replication: improve MRF healing. (#15556)

This PR improves the replication failure healing by persisting
most recent failures to disk and re-queuing them until the replication
is successful.

While this does not eliminate the need for healing during a full scan, 
queuing MRF vastly improves the ETA to keeping replicated buckets 
in sync as it does not wait for the scanner visit to detect unreplicated 
object versions.

.github/workflows/depsreview.yaml

@@ -0,0 +1,14 @@
+name: 'Dependency Review'
+on: [pull_request]
+
+permissions:
+  contents: read
+
+jobs:
+  dependency-review:
+    runs-on: ubuntu-latest
+    steps:
+      - name: 'Checkout Repository'
+        uses: actions/checkout@v3
+      - name: 'Dependency Review'
+        uses: actions/dependency-review-action@v1

.github/workflows/go-cross.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x, 1.18.x]
+        go-version: [1.18.x]
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2

.github/workflows/go-fips.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.11b7, 1.18.3b7]
+        go-version: [1.18.5b7]
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@v2

.github/workflows/go-healing.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.18.x]
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@v2
@@ -48,3 +48,5 @@ jobs:
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make verify-healing
           make verify-healing-inconsistent-versions
+          make verify-healing-with-root-disks
+          make verify-healing-with-rewrite

.github/workflows/go-lint.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x, 1.18.x]
+        go-version: [1.18.x]
         os: [ubuntu-latest, windows-latest]
     steps:
       - uses: actions/checkout@v2

.github/workflows/go.yml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x, 1.18.x]
+        go-version: [1.18.x]
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@v2

.github/workflows/iam-integrations.yaml

@@ -61,7 +61,7 @@ jobs:
       # are turned off - i.e. if ldap="", then ldap server is not enabled for
       # the tests.
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.18.x]
         ldap: ["", "localhost:389"]
         etcd: ["", "http://localhost:2379"]
         openid: ["", "http://127.0.0.1:5556/dex"]

.github/workflows/markdown-lint.yaml

@@ -27,4 +27,4 @@ jobs:
         npm install -g markdownlint-cli
         markdownlint --fix '**/*.md' \
           --config /home/runner/work/minio/minio/.github/markdown-lint-cfg.yaml \
-          --disable MD013 MD040
+          --disable MD013 MD040 MD051

.github/workflows/replication.yaml

@@ -1,4 +1,4 @@
-name: Multi-site replication tests
+name: MinIO advanced tests
 
 on:
   pull_request:
@@ -16,12 +16,12 @@ permissions:
 
 jobs:
   replication-test:
-    name: Replication Tests with Go ${{ matrix.go-version }}
+    name: Advanced Tests with Go ${{ matrix.go-version }}
     runs-on: ubuntu-latest
 
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.18.x]
 
     steps:
       - uses: actions/checkout@v2
@@ -37,11 +37,18 @@ jobs:
           key: ${{ runner.os }}-${{ matrix.go-version }}-go-${{ hashFiles('**/go.sum') }}
           restore-keys: |
             ${{ runner.os }}-${{ matrix.go-version }}-go-
+      - name: Test Decom
+        run: |
+          sudo sysctl net.ipv6.conf.all.disable_ipv6=0
+          sudo sysctl net.ipv6.conf.default.disable_ipv6=0
+          make test-decom
+
       - name: Test Replication
         run: |
           sudo sysctl net.ipv6.conf.all.disable_ipv6=0
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make test-replication
+
       - name: Test MinIO IDP for automatic site replication
         run: |
           sudo sysctl net.ipv6.conf.all.disable_ipv6=0

.github/workflows/upgrade-ci-cd.yaml

@@ -20,7 +20,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.18.x]
         os: [ubuntu-latest]
 
     steps:

.gitignore

@@ -9,8 +9,7 @@ site/
 /.idea/
 /Minio.iml
 **/access.log
-vendor/**/*.js
-vendor/**/*.json
+vendor/
 .DS_Store
 *.syso
 coverage.txt
@@ -32,4 +31,10 @@ hash-set
 minio.RELEASE*
 mc
 nancy
-inspects/*
+inspects/*
+docs/debugging/s3-verify/s3-verify
+docs/debugging/xl-meta/xl-meta
+docs/debugging/s3-check-md5/s3-check-md5
+docs/debugging/hash-set/hash-set
+docs/debugging/healing-bin/healing-bin
+docs/debugging/inspect/inspect

.golangci.yml

@@ -14,13 +14,9 @@ linters:
     - govet
     - revive
     - ineffassign
-    - gosimple
     - deadcode
-    - structcheck
     - gomodguard
     - gofmt
-    - unused
-    - structcheck
     - unconvert
     - varcheck
     - gocritic

.nancy-ignore

@@ -2,3 +2,5 @@ CVE-2020-26160
 CVE-2020-15136
 CVE-2020-15115
 CVE-2020-15114
+CVE-2022-2835
+CVE-2022-2837

CREDITS

@@ -2741,186 +2741,6 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 ================================================================
 
-github.com/briandowns/spinner
-https://github.com/briandowns/spinner
-----------------------------------------------------------------
-                                 Apache License
-                           Version 2.0, January 2004
-                        http://www.apache.org/licenses/
-
-   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
-
-   1. Definitions.
-
-      "License" shall mean the terms and conditions for use, reproduction,
-      and distribution as defined by Sections 1 through 9 of this document.
-
-      "Licensor" shall mean the copyright owner or entity authorized by
-      the copyright owner that is granting the License.
-
-      "Legal Entity" shall mean the union of the acting entity and all
-      other entities that control, are controlled by, or are under common
-      control with that entity. For the purposes of this definition,
-      "control" means (i) the power, direct or indirect, to cause the
-      direction or management of such entity, whether by contract or
-      otherwise, or (ii) ownership of fifty percent (50%) or more of the
-      outstanding shares, or (iii) beneficial ownership of such entity.
-
-      "You" (or "Your") shall mean an individual or Legal Entity
-      exercising permissions granted by this License.
-
-      "Source" form shall mean the preferred form for making modifications,
-      including but not limited to software source code, documentation
-      source, and configuration files.
-
-      "Object" form shall mean any form resulting from mechanical
-      transformation or translation of a Source form, including but
-      not limited to compiled object code, generated documentation,
-      and conversions to other media types.
-
-      "Work" shall mean the work of authorship, whether in Source or
-      Object form, made available under the License, as indicated by a
-      copyright notice that is included in or attached to the work
-      (an example is provided in the Appendix below).
-
-      "Derivative Works" shall mean any work, whether in Source or Object
-      form, that is based on (or derived from) the Work and for which the
-      editorial revisions, annotations, elaborations, or other modifications
-      represent, as a whole, an original work of authorship. For the purposes
-      of this License, Derivative Works shall not include works that remain
-      separable from, or merely link (or bind by name) to the interfaces of,
-      the Work and Derivative Works thereof.
-
-      "Contribution" shall mean any work of authorship, including
-      the original version of the Work and any modifications or additions
-      to that Work or Derivative Works thereof, that is intentionally
-      submitted to Licensor for inclusion in the Work by the copyright owner
-      or by an individual or Legal Entity authorized to submit on behalf of
-      the copyright owner. For the purposes of this definition, "submitted"
-      means any form of electronic, verbal, or written communication sent
-      to the Licensor or its representatives, including but not limited to
-      communication on electronic mailing lists, source code control systems,
-      and issue tracking systems that are managed by, or on behalf of, the
-      Licensor for the purpose of discussing and improving the Work, but
-      excluding communication that is conspicuously marked or otherwise
-      designated in writing by the copyright owner as "Not a Contribution."
-
-      "Contributor" shall mean Licensor and any individual or Legal Entity
-      on behalf of whom a Contribution has been received by Licensor and
-      subsequently incorporated within the Work.
-
-   2. Grant of Copyright License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      copyright license to reproduce, prepare Derivative Works of,
-      publicly display, publicly perform, sublicense, and distribute the
-      Work and such Derivative Works in Source or Object form.
-
-   3. Grant of Patent License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      (except as stated in this section) patent license to make, have made,
-      use, offer to sell, sell, import, and otherwise transfer the Work,
-      where such license applies only to those patent claims licensable
-      by such Contributor that are necessarily infringed by their
-      Contribution(s) alone or by combination of their Contribution(s)
-      with the Work to which such Contribution(s) was submitted. If You
-      institute patent litigation against any entity (including a
-      cross-claim or counterclaim in a lawsuit) alleging that the Work
-      or a Contribution incorporated within the Work constitutes direct
-      or contributory patent infringement, then any patent licenses
-      granted to You under this License for that Work shall terminate
-      as of the date such litigation is filed.
-
-   4. Redistribution. You may reproduce and distribute copies of the
-      Work or Derivative Works thereof in any medium, with or without
-      modifications, and in Source or Object form, provided that You
-      meet the following conditions:
-
-      (a) You must give any other recipients of the Work or
-          Derivative Works a copy of this License; and
-
-      (b) You must cause any modified files to carry prominent notices
-          stating that You changed the files; and
-
-      (c) You must retain, in the Source form of any Derivative Works
-          that You distribute, all copyright, patent, trademark, and
-          attribution notices from the Source form of the Work,
-          excluding those notices that do not pertain to any part of
-          the Derivative Works; and
-
-      (d) If the Work includes a "NOTICE" text file as part of its
-          distribution, then any Derivative Works that You distribute must
-          include a readable copy of the attribution notices contained
-          within such NOTICE file, excluding those notices that do not
-          pertain to any part of the Derivative Works, in at least one
-          of the following places: within a NOTICE text file distributed
-          as part of the Derivative Works; within the Source form or
-          documentation, if provided along with the Derivative Works; or,
-          within a display generated by the Derivative Works, if and
-          wherever such third-party notices normally appear. The contents
-          of the NOTICE file are for informational purposes only and
-          do not modify the License. You may add Your own attribution
-          notices within Derivative Works that You distribute, alongside
-          or as an addendum to the NOTICE text from the Work, provided
-          that such additional attribution notices cannot be construed
-          as modifying the License.
-
-      You may add Your own copyright statement to Your modifications and
-      may provide additional or different license terms and conditions
-      for use, reproduction, or distribution of Your modifications, or
-      for any such Derivative Works as a whole, provided Your use,
-      reproduction, and distribution of the Work otherwise complies with
-      the conditions stated in this License.
-
-   5. Submission of Contributions. Unless You explicitly state otherwise,
-      any Contribution intentionally submitted for inclusion in the Work
-      by You to the Licensor shall be under the terms and conditions of
-      this License, without any additional terms or conditions.
-      Notwithstanding the above, nothing herein shall supersede or modify
-      the terms of any separate license agreement you may have executed
-      with Licensor regarding such Contributions.
-
-   6. Trademarks. This License does not grant permission to use the trade
-      names, trademarks, service marks, or product names of the Licensor,
-      except as required for reasonable and customary use in describing the
-      origin of the Work and reproducing the content of the NOTICE file.
-
-   7. Disclaimer of Warranty. Unless required by applicable law or
-      agreed to in writing, Licensor provides the Work (and each
-      Contributor provides its Contributions) on an "AS IS" BASIS,
-      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
-      implied, including, without limitation, any warranties or conditions
-      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
-      PARTICULAR PURPOSE. You are solely responsible for determining the
-      appropriateness of using or redistributing the Work and assume any
-      risks associated with Your exercise of permissions under this License.
-
-   8. Limitation of Liability. In no event and under no legal theory,
-      whether in tort (including negligence), contract, or otherwise,
-      unless required by applicable law (such as deliberate and grossly
-      negligent acts) or agreed to in writing, shall any Contributor be
-      liable to You for damages, including any direct, indirect, special,
-      incidental, or consequential damages of any character arising as a
-      result of this License or out of the use or inability to use the
-      Work (including but not limited to damages for loss of goodwill,
-      work stoppage, computer failure or malfunction, or any and all
-      other commercial damages or losses), even if such Contributor
-      has been advised of the possibility of such damages.
-
-   9. Accepting Warranty or Additional Liability. While redistributing
-      the Work or Derivative Works thereof, You may choose to offer,
-      and charge a fee for, acceptance of support, warranty, indemnity,
-      or other liability obligations and/or rights consistent with this
-      License. However, in accepting such obligations, You may act only
-      on Your own behalf and on Your sole responsibility, not on behalf
-      of any other Contributor, and only if You agree to indemnify,
-      defend, and hold each Contributor harmless for any liability
-      incurred by, or claims asserted against, such Contributor by reason
-      of your accepting any such warranty or additional liability.
-
-================================================================
-
 github.com/buger/jsonparser
 https://github.com/buger/jsonparser
 ----------------------------------------------------------------
@@ -4885,26 +4705,300 @@ SOFTWARE.
 github.com/eclipse/paho.mqtt.golang
 https://github.com/eclipse/paho.mqtt.golang
 ----------------------------------------------------------------
-This project is dual licensed under the Eclipse Public License 1.0 and the
-Eclipse Distribution License 1.0 as described in the epl-v10 and edl-v10 files.
+Eclipse Public License - v 2.0 (EPL-2.0)
 
-The EDL is copied below in order to pass the pkg.go.dev license check (https://pkg.go.dev/license-policy).
+This program and the accompanying materials
+are made available under the terms of the Eclipse Public License v2.0
+and Eclipse Distribution License v1.0 which accompany this distribution.
+
+The Eclipse Public License is available at
+  https://www.eclipse.org/legal/epl-2.0/
+and the Eclipse Distribution License is available at
+  http://www.eclipse.org/org/documents/edl-v10.php.
+
+For an explanation of what dual-licensing means to you, see:
+https://www.eclipse.org/legal/eplfaq.php#DUALLIC
 
 ****
-Eclipse Distribution License - v 1.0
+The epl-2.0 is copied below in order to pass the pkg.go.dev license check (https://pkg.go.dev/license-policy).
+****
+Eclipse Public License - v 2.0
 
-Copyright (c) 2007, Eclipse Foundation, Inc. and its licensors.
+    THE ACCOMPANYING PROGRAM IS PROVIDED UNDER THE TERMS OF THIS ECLIPSE
+    PUBLIC LICENSE ("AGREEMENT"). ANY USE, REPRODUCTION OR DISTRIBUTION
+    OF THE PROGRAM CONSTITUTES RECIPIENT'S ACCEPTANCE OF THIS AGREEMENT.
 
-All rights reserved.
+1. DEFINITIONS
 
-Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+"Contribution" means:
 
-    Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
-    Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
-    Neither the name of the Eclipse Foundation, Inc. nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
+  a) in the case of the initial Contributor, the initial content
+     Distributed under this Agreement, and
 
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+  b) in the case of each subsequent Contributor:
+     i) changes to the Program, and
+     ii) additions to the Program;
+  where such changes and/or additions to the Program originate from
+  and are Distributed by that particular Contributor. A Contribution
+  "originates" from a Contributor if it was added to the Program by
+  such Contributor itself or anyone acting on such Contributor's behalf.
+  Contributions do not include changes or additions to the Program that
+  are not Modified Works.
 
+"Contributor" means any person or entity that Distributes the Program.
+
+"Licensed Patents" mean patent claims licensable by a Contributor which
+are necessarily infringed by the use or sale of its Contribution alone
+or when combined with the Program.
+
+"Program" means the Contributions Distributed in accordance with this
+Agreement.
+
+"Recipient" means anyone who receives the Program under this Agreement
+or any Secondary License (as applicable), including Contributors.
+
+"Derivative Works" shall mean any work, whether in Source Code or other
+form, that is based on (or derived from) the Program and for which the
+editorial revisions, annotations, elaborations, or other modifications
+represent, as a whole, an original work of authorship.
+
+"Modified Works" shall mean any work in Source Code or other form that
+results from an addition to, deletion from, or modification of the
+contents of the Program, including, for purposes of clarity any new file
+in Source Code form that contains any contents of the Program. Modified
+Works shall not include works that contain only declarations,
+interfaces, types, classes, structures, or files of the Program solely
+in each case in order to link to, bind by name, or subclass the Program
+or Modified Works thereof.
+
+"Distribute" means the acts of a) distributing or b) making available
+in any manner that enables the transfer of a copy.
+
+"Source Code" means the form of a Program preferred for making
+modifications, including but not limited to software source code,
+documentation source, and configuration files.
+
+"Secondary License" means either the GNU General Public License,
+Version 2.0, or any later versions of that license, including any
+exceptions or additional permissions as identified by the initial
+Contributor.
+
+2. GRANT OF RIGHTS
+
+  a) Subject to the terms of this Agreement, each Contributor hereby
+  grants Recipient a non-exclusive, worldwide, royalty-free copyright
+  license to reproduce, prepare Derivative Works of, publicly display,
+  publicly perform, Distribute and sublicense the Contribution of such
+  Contributor, if any, and such Derivative Works.
+
+  b) Subject to the terms of this Agreement, each Contributor hereby
+  grants Recipient a non-exclusive, worldwide, royalty-free patent
+  license under Licensed Patents to make, use, sell, offer to sell,
+  import and otherwise transfer the Contribution of such Contributor,
+  if any, in Source Code or other form. This patent license shall
+  apply to the combination of the Contribution and the Program if, at
+  the time the Contribution is added by the Contributor, such addition
+  of the Contribution causes such combination to be covered by the
+  Licensed Patents. The patent license shall not apply to any other
+  combinations which include the Contribution. No hardware per se is
+  licensed hereunder.
+
+  c) Recipient understands that although each Contributor grants the
+  licenses to its Contributions set forth herein, no assurances are
+  provided by any Contributor that the Program does not infringe the
+  patent or other intellectual property rights of any other entity.
+  Each Contributor disclaims any liability to Recipient for claims
+  brought by any other entity based on infringement of intellectual
+  property rights or otherwise. As a condition to exercising the
+  rights and licenses granted hereunder, each Recipient hereby
+  assumes sole responsibility to secure any other intellectual
+  property rights needed, if any. For example, if a third party
+  patent license is required to allow Recipient to Distribute the
+  Program, it is Recipient's responsibility to acquire that license
+  before distributing the Program.
+
+  d) Each Contributor represents that to its knowledge it has
+  sufficient copyright rights in its Contribution, if any, to grant
+  the copyright license set forth in this Agreement.
+
+  e) Notwithstanding the terms of any Secondary License, no
+  Contributor makes additional grants to any Recipient (other than
+  those set forth in this Agreement) as a result of such Recipient's
+  receipt of the Program under the terms of a Secondary License
+  (if permitted under the terms of Section 3).
+
+3. REQUIREMENTS
+
+3.1 If a Contributor Distributes the Program in any form, then:
+
+  a) the Program must also be made available as Source Code, in
+  accordance with section 3.2, and the Contributor must accompany
+  the Program with a statement that the Source Code for the Program
+  is available under this Agreement, and informs Recipients how to
+  obtain it in a reasonable manner on or through a medium customarily
+  used for software exchange; and
+
+  b) the Contributor may Distribute the Program under a license
+  different than this Agreement, provided that such license:
+     i) effectively disclaims on behalf of all other Contributors all
+     warranties and conditions, express and implied, including
+     warranties or conditions of title and non-infringement, and
+     implied warranties or conditions of merchantability and fitness
+     for a particular purpose;
+
+     ii) effectively excludes on behalf of all other Contributors all
+     liability for damages, including direct, indirect, special,
+     incidental and consequential damages, such as lost profits;
+
+     iii) does not attempt to limit or alter the recipients' rights
+     in the Source Code under section 3.2; and
+
+     iv) requires any subsequent distribution of the Program by any
+     party to be under a license that satisfies the requirements
+     of this section 3.
+
+3.2 When the Program is Distributed as Source Code:
+
+  a) it must be made available under this Agreement, or if the
+  Program (i) is combined with other material in a separate file or
+  files made available under a Secondary License, and (ii) the initial
+  Contributor attached to the Source Code the notice described in
+  Exhibit A of this Agreement, then the Program may be made available
+  under the terms of such Secondary Licenses, and
+
+  b) a copy of this Agreement must be included with each copy of
+  the Program.
+
+3.3 Contributors may not remove or alter any copyright, patent,
+trademark, attribution notices, disclaimers of warranty, or limitations
+of liability ("notices") contained within the Program from any copy of
+the Program which they Distribute, provided that Contributors may add
+their own appropriate notices.
+
+4. COMMERCIAL DISTRIBUTION
+
+Commercial distributors of software may accept certain responsibilities
+with respect to end users, business partners and the like. While this
+license is intended to facilitate the commercial use of the Program,
+the Contributor who includes the Program in a commercial product
+offering should do so in a manner which does not create potential
+liability for other Contributors. Therefore, if a Contributor includes
+the Program in a commercial product offering, such Contributor
+("Commercial Contributor") hereby agrees to defend and indemnify every
+other Contributor ("Indemnified Contributor") against any losses,
+damages and costs (collectively "Losses") arising from claims, lawsuits
+and other legal actions brought by a third party against the Indemnified
+Contributor to the extent caused by the acts or omissions of such
+Commercial Contributor in connection with its distribution of the Program
+in a commercial product offering. The obligations in this section do not
+apply to any claims or Losses relating to any actual or alleged
+intellectual property infringement. In order to qualify, an Indemnified
+Contributor must: a) promptly notify the Commercial Contributor in
+writing of such claim, and b) allow the Commercial Contributor to control,
+and cooperate with the Commercial Contributor in, the defense and any
+related settlement negotiations. The Indemnified Contributor may
+participate in any such claim at its own expense.
+
+For example, a Contributor might include the Program in a commercial
+product offering, Product X. That Contributor is then a Commercial
+Contributor. If that Commercial Contributor then makes performance
+claims, or offers warranties related to Product X, those performance
+claims and warranties are such Commercial Contributor's responsibility
+alone. Under this section, the Commercial Contributor would have to
+defend claims against the other Contributors related to those performance
+claims and warranties, and if a court requires any other Contributor to
+pay any damages as a result, the Commercial Contributor must pay
+those damages.
+
+5. NO WARRANTY
+
+EXCEPT AS EXPRESSLY SET FORTH IN THIS AGREEMENT, AND TO THE EXTENT
+PERMITTED BY APPLICABLE LAW, THE PROGRAM IS PROVIDED ON AN "AS IS"
+BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, EITHER EXPRESS OR
+IMPLIED INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OR CONDITIONS OF
+TITLE, NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR
+PURPOSE. Each Recipient is solely responsible for determining the
+appropriateness of using and distributing the Program and assumes all
+risks associated with its exercise of rights under this Agreement,
+including but not limited to the risks and costs of program errors,
+compliance with applicable laws, damage to or loss of data, programs
+or equipment, and unavailability or interruption of operations.
+
+6. DISCLAIMER OF LIABILITY
+
+EXCEPT AS EXPRESSLY SET FORTH IN THIS AGREEMENT, AND TO THE EXTENT
+PERMITTED BY APPLICABLE LAW, NEITHER RECIPIENT NOR ANY CONTRIBUTORS
+SHALL HAVE ANY LIABILITY FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING WITHOUT LIMITATION LOST
+PROFITS), HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OR DISTRIBUTION OF THE PROGRAM OR THE
+EXERCISE OF ANY RIGHTS GRANTED HEREUNDER, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGES.
+
+7. GENERAL
+
+If any provision of this Agreement is invalid or unenforceable under
+applicable law, it shall not affect the validity or enforceability of
+the remainder of the terms of this Agreement, and without further
+action by the parties hereto, such provision shall be reformed to the
+minimum extent necessary to make such provision valid and enforceable.
+
+If Recipient institutes patent litigation against any entity
+(including a cross-claim or counterclaim in a lawsuit) alleging that the
+Program itself (excluding combinations of the Program with other software
+or hardware) infringes such Recipient's patent(s), then such Recipient's
+rights granted under Section 2(b) shall terminate as of the date such
+litigation is filed.
+
+All Recipient's rights under this Agreement shall terminate if it
+fails to comply with any of the material terms or conditions of this
+Agreement and does not cure such failure in a reasonable period of
+time after becoming aware of such noncompliance. If all Recipient's
+rights under this Agreement terminate, Recipient agrees to cease use
+and distribution of the Program as soon as reasonably practicable.
+However, Recipient's obligations under this Agreement and any licenses
+granted by Recipient relating to the Program shall continue and survive.
+
+Everyone is permitted to copy and distribute copies of this Agreement,
+but in order to avoid inconsistency the Agreement is copyrighted and
+may only be modified in the following manner. The Agreement Steward
+reserves the right to publish new versions (including revisions) of
+this Agreement from time to time. No one other than the Agreement
+Steward has the right to modify this Agreement. The Eclipse Foundation
+is the initial Agreement Steward. The Eclipse Foundation may assign the
+responsibility to serve as the Agreement Steward to a suitable separate
+entity. Each new version of the Agreement will be given a distinguishing
+version number. The Program (including Contributions) may always be
+Distributed subject to the version of the Agreement under which it was
+received. In addition, after a new version of the Agreement is published,
+Contributor may elect to Distribute the Program (including its
+Contributions) under the new version.
+
+Except as expressly stated in Sections 2(a) and 2(b) above, Recipient
+receives no rights or licenses to the intellectual property of any
+Contributor under this Agreement, whether expressly, by implication,
+estoppel or otherwise. All rights in the Program not expressly granted
+under this Agreement are reserved. Nothing in this Agreement is intended
+to be enforceable by any entity that is not a Contributor or Recipient.
+No third-party beneficiary rights are created under this Agreement.
+
+Exhibit A - Form of Secondary Licenses Notice
+
+"This Source Code may also be made available under the following
+Secondary Licenses when the conditions for such availability set forth
+in the Eclipse Public License, v. 2.0 are satisfied: {name license(s),
+version(s), and exceptions or additional permissions here}."
+
+  Simply including a copy of this Agreement, including this Exhibit A
+  is not sufficient to license the Source Code under Secondary Licenses.
+
+  If it is not possible or desirable to put the notice in a particular
+  file, then You may include the notice in a location (such as a LICENSE
+  file in a relevant directory) where a recipient would be likely to
+  look for such a notice.
+
+  You may add additional accurate notices of copyright ownership.
 
 ================================================================
 
@@ -22029,6 +22123,40 @@ OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 
+================================================================
+
+github.com/pierrec/lz4/v4
+https://github.com/pierrec/lz4/v4
+----------------------------------------------------------------
+Copyright (c) 2015, Pierre Curto
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+* Redistributions of source code must retain the above copyright notice, this
+  list of conditions and the following disclaimer.
+
+* Redistributions in binary form must reproduce the above copyright notice,
+  this list of conditions and the following disclaimer in the documentation
+  and/or other materials provided with the distribution.
+
+* Neither the name of xxHash nor the names of its
+  contributors may be used to endorse or promote products derived from
+  this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+
 ================================================================
 
 github.com/pkg/errors
@@ -23215,6 +23343,213 @@ https://github.com/prometheus/procfs
 
 ================================================================
 
+github.com/prometheus/prom2json
+https://github.com/prometheus/prom2json
+----------------------------------------------------------------
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.
+
+================================================================
+
 github.com/rcrowley/go-metrics
 https://github.com/rcrowley/go-metrics
 ----------------------------------------------------------------
@@ -27415,17 +27750,3 @@ limitations under the License.
 
 ================================================================
 
-maze.io/x/duration
-https://maze.io/x/duration
-----------------------------------------------------------------
-MIT License
-Copyright (c) <year> <copyright holders>
-
-Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-================================================================
-

Dockerfile.hotfix

@@ -1,4 +1,4 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.5
+FROM registry.access.redhat.com/ubi8/ubi-minimal:8.6
 
 ARG RELEASE
 

Dockerfile.release

@@ -1,4 +1,4 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.5
+FROM registry.access.redhat.com/ubi8/ubi-minimal:8.6
 
 ARG TARGETARCH
 

Dockerfile.release.fips

@@ -1,4 +1,4 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.5
+FROM registry.access.redhat.com/ubi8/ubi-minimal:8.6
 
 ARG TARGETARCH
 

Makefile

@@ -15,7 +15,7 @@ checks: ## check dependencies
 	@(env bash $(PWD)/buildscripts/checkdeps.sh)
 
 help: ## print this help
-	@grep -E '^[a-zA-Z_-]+:.*?## .*$$' Makefile | awk 'BEGIN {FS = ":.*?## "}; {printf "\033[36m%-30s\033[0m %s\n", $$1, $$2}'
+	@grep -E '^[a-zA-Z_-]+:.*?## .*$$' Makefile | awk 'BEGIN {FS = ":.*?## "}; {printf "\033[36m%-40s\033[0m %s\n", $$1, $$2}'
 
 getdeps: ## fetch necessary dependencies
 	@mkdir -p ${GOPATH}/bin
@@ -39,7 +39,14 @@ lint: ## runs golangci-lint suite of linters
 check: test
 test: verifiers build ## builds minio, runs linters, tests
 	@echo "Running unit tests"
-	@CGO_ENABLED=0 go test -tags kqueue ./...
+	@MINIO_API_REQUESTS_MAX=10000 CGO_ENABLED=0 go test -tags kqueue ./...
+
+test-decom: install
+	@echo "Running minio decom tests"
+	@env bash $(PWD)/docs/distributed/decom.sh
+	@env bash $(PWD)/docs/distributed/decom-encrypted.sh
+	@env bash $(PWD)/docs/distributed/decom-encrypted-sse-s3.sh
+	@env bash $(PWD)/docs/distributed/decom-compressed-sse-s3.sh
 
 test-upgrade: build
 	@echo "Running minio upgrade tests"
@@ -51,9 +58,9 @@ test-race: verifiers build ## builds minio, runs linters, tests (race)
 
 test-iam: build ## verify IAM (external IDP, etcd backends)
 	@echo "Running tests for IAM (external IDP, etcd backends)"
-	@CGO_ENABLED=0 go test -tags kqueue -v -run TestIAM* ./cmd
+	@MINIO_API_REQUESTS_MAX=10000 CGO_ENABLED=0 go test -tags kqueue -v -run TestIAM* ./cmd
 	@echo "Running tests for IAM (external IDP, etcd backends) with -race"
-	@GORACE=history_size=7 CGO_ENABLED=1 go test -race -tags kqueue -v -run TestIAM* ./cmd
+	@MINIO_API_REQUESTS_MAX=10000 GORACE=history_size=7 CGO_ENABLED=1 go test -race -tags kqueue -v -run TestIAM* ./cmd
 
 test-replication: install ## verify multi site replication
 	@echo "Running tests for replicating three sites"
@@ -83,6 +90,16 @@ verify-healing: ## verify healing and replacing disks with minio binary
 	@(env bash $(PWD)/buildscripts/verify-healing.sh)
 	@(env bash $(PWD)/buildscripts/unaligned-healing.sh)
 
+verify-healing-with-root-disks: ## verify healing root disks
+	@echo "Verify healing with root drives"
+	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
+	@(env bash $(PWD)/buildscripts/verify-healing-with-root-disks.sh)
+
+verify-healing-with-rewrite: ## verify healing to rewrite old xl.meta -> new xl.meta
+	@echo "Verify healing with rewrite"
+	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
+	@(env bash $(PWD)/buildscripts/rewrite-old-new.sh)
+
 verify-healing-inconsistent-versions: ## verify resolving inconsistent versions
 	@echo "Verify resolving inconsistent versions build with race"
 	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
@@ -128,6 +145,8 @@ clean: ## cleanup all generated assets
 	@echo "Cleaning up all the generated files"
 	@find . -name '*.test' | xargs rm -fv
 	@find . -name '*~' | xargs rm -fv
+	@find . -name '.#*#' | xargs rm -fv
+	@find . -name '#*#' | xargs rm -fv
 	@rm -rvf minio
 	@rm -rvf build
 	@rm -rvf release

README.fips.md

@@ -2,6 +2,6 @@
 
 MinIO creates FIPS builds using a patched version of the Go compiler (that uses BoringCrypto, from BoringSSL, which is [FIPS 140-2 validated](https://csrc.nist.gov/csrc/media/projects/cryptographic-module-validation-program/documents/security-policies/140sp2964.pdf)) published by the Golang Team [here](https://github.com/golang/go/tree/dev.boringcrypto/misc/boring).
 
-MinIO FIPS executables are available at http://dl.min.io - they are only published for `linux-amd64` architecture as binary files with the suffix `.fips`. We also publish corresponding container images to our official image repositories.
+MinIO FIPS executables are available at <http://dl.min.io> - they are only published for `linux-amd64` architecture as binary files with the suffix `.fips`. We also publish corresponding container images to our official image repositories.
 
 We are not making any statements or representations about the suitability of this code or build in relation to the FIPS 140-2 standard. Interested users will have to evaluate for themselves whether this is useful for their own purposes.

README.md

@@ -125,7 +125,7 @@ You can also connect using any S3-compatible tool, such as the MinIO Client `mc`
 
 ## Install from Source
 
-Use the following commands to compile and run a standalone MinIO server from source. Source installation is only intended for developers and advanced users. If you do not have a working Golang environment, please follow [How to install Golang](https://golang.org/doc/install). Minimum version required is [go1.17](https://golang.org/dl/#stable)
+Use the following commands to compile and run a standalone MinIO server from source. Source installation is only intended for developers and advanced users. If you do not have a working Golang environment, please follow [How to install Golang](https://golang.org/doc/install). Minimum version required is [go1.18](https://golang.org/dl/#stable)
 
 ```sh
 GO111MODULE=on go install github.com/minio/minio@latest
@@ -236,16 +236,16 @@ Upgrades require zero downtime in MinIO, all upgrades are non-disruptive, all tr
 mc admin update <minio alias, e.g., myminio>
 ```
 
-- For deployments without external internet access (e.g. airgapped environments), download the binary from <https://dl.min.io> and replace the existing MinIO binary let's say for example `/opt/bin/minio`, apply executable permissions `chmod +x /opt/bin/minio` and do `mc admin service restart alias/`.
+- For deployments without external internet access (e.g. airgapped environments), download the binary from <https://dl.min.io> and replace the existing MinIO binary let's say for example `/opt/bin/minio`, apply executable permissions `chmod +x /opt/bin/minio` and proceed to perform `mc admin service restart alias/`.
 
-- For installations using Systemd MinIO service, upgrade via RPM/DEB packages **parallelly** on all servers or replace the binary lets say `/opt/bin/minio` on all nodes, apply executable permissions `chmod +x /opt/bin/minio`. Proceed to perform `systemctl restart minio` across all nodes in **parallel**.
+- For installations using Systemd MinIO service, upgrade via RPM/DEB packages **parallelly** on all servers or replace the binary lets say `/opt/bin/minio` on all nodes, apply executable permissions `chmod +x /opt/bin/minio` and process to perform `mc admin service restart alias/`.
 
 ### Upgrade Checklist
 
 - Test all upgrades in a lower environment (DEV, QA, UAT) before applying to production. Performing blind upgrades in production environments carries significant risk.
-- Read the release notes for the targeted MinIO release *before* performing any installation, there is no forced requirement to upgrade to latest releases every week. If it has a bug fix you are looking for then yes, else avoid actively upgrading a running production system.
-- If you plan to use `mc admin update`, MinIO process must have write access to the parent directory to provide in-place upgrades.
-- `mc admin update` is not supported in kubernetes/container environments, container environments provide their own mechanisms for container updates.
+- Read the release notes for MinIO *before* performing any upgrade, there is no forced requirement to upgrade to latest releases upon every releases. Some releases may not be relevant to your setup, avoid upgrading production environments unnecessarily.
+- If you plan to use `mc admin update`, MinIO process must have write access to the parent directory where the binary is present on the host system.
+- `mc admin update` is not supported and should be avoided in kubernetes/container environments, please upgrade containers by upgrading relevant container images.
 - **We do not recommend upgrading one MinIO server at a time, the product is designed to support parallel upgrades please follow our recommended guidelines.**
 
 ## Explore Further

buildscripts/gen-ldflags.go

@@ -24,14 +24,18 @@ import (
 	"fmt"
 	"os"
 	"os/exec"
+	"strconv"
 	"strings"
 	"time"
 )
 
 func genLDFlags(version string) string {
+	releaseTag, date := releaseTag(version)
+	copyrightYear := strconv.Itoa(date.Year())
 	ldflagsStr := "-s -w"
 	ldflagsStr += " -X github.com/minio/minio/cmd.Version=" + version
-	ldflagsStr += " -X github.com/minio/minio/cmd.ReleaseTag=" + releaseTag(version)
+	ldflagsStr += " -X github.com/minio/minio/cmd.CopyrightYear=" + copyrightYear
+	ldflagsStr += " -X github.com/minio/minio/cmd.ReleaseTag=" + releaseTag
 	ldflagsStr += " -X github.com/minio/minio/cmd.CommitID=" + commitID()
 	ldflagsStr += " -X github.com/minio/minio/cmd.ShortCommitID=" + commitID()[:12]
 	ldflagsStr += " -X github.com/minio/minio/cmd.GOPATH=" + os.Getenv("GOPATH")
@@ -40,7 +44,7 @@ func genLDFlags(version string) string {
 }
 
 // genReleaseTag prints release tag to the console for easy git tagging.
-func releaseTag(version string) string {
+func releaseTag(version string) (string, time.Time) {
 	relPrefix := "DEVELOPMENT"
 	if prefix := os.Getenv("MINIO_RELEASE"); prefix != "" {
 		relPrefix = prefix
@@ -53,14 +57,17 @@ func releaseTag(version string) string {
 
 	relTag := strings.Replace(version, " ", "-", -1)
 	relTag = strings.Replace(relTag, ":", "-", -1)
+	t, err := time.Parse("2006-01-02T15-04-05Z", relTag)
+	if err != nil {
+		panic(err)
+	}
 	relTag = strings.Replace(relTag, ",", "", -1)
 	relTag = relPrefix + "." + relTag
-
 	if relSuffix != "" {
 		relTag += "." + relSuffix
 	}
 
-	return relTag
+	return relTag, t
 }
 
 // commitID returns the abbreviated commit-id hash of the last commit.

buildscripts/heal-manual.go

@@ -0,0 +1,87 @@
+//go:build ignore
+// +build ignore
+
+//
+// MinIO Object Storage (c) 2022 MinIO, Inc.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//      http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+//
+
+package main
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"log"
+	"os"
+	"time"
+
+	"github.com/minio/madmin-go"
+)
+
+func main() {
+	// Note: YOUR-ACCESSKEYID, YOUR-SECRETACCESSKEY are
+	// dummy values, please replace them with original values.
+
+	// API requests are secure (HTTPS) if secure=true and insecure (HTTP) otherwise.
+	// New returns an MinIO Admin client object.
+	madmClnt, err := madmin.New(os.Args[1], os.Args[2], os.Args[3], false)
+	if err != nil {
+		log.Fatalln(err)
+	}
+
+	opts := madmin.HealOpts{
+		Recursive: true,                  // recursively heal all objects at 'prefix'
+		Remove:    true,                  // remove content that has lost quorum and not recoverable
+		Recreate:  true,                  // rewrite all old non-inlined xl.meta to new xl.meta
+		ScanMode:  madmin.HealNormalScan, // by default do not do 'deep' scanning
+	}
+
+	start, _, err := madmClnt.Heal(context.Background(), "healing-rewrite-bucket", "", opts, "", false, false)
+	if err != nil {
+		log.Fatalln(err)
+	}
+	fmt.Println("Healstart sequence ===")
+	enc := json.NewEncoder(os.Stdout)
+	if err = enc.Encode(&start); err != nil {
+		log.Fatalln(err)
+	}
+
+	fmt.Println()
+	for {
+		_, status, err := madmClnt.Heal(context.Background(), "healing-rewrite-bucket", "", opts, start.ClientToken, false, false)
+		if status.Summary == "finished" {
+			fmt.Println("Healstatus on items ===")
+			for _, item := range status.Items {
+				if err = enc.Encode(&item); err != nil {
+					log.Fatalln(err)
+				}
+			}
+			break
+		}
+		if status.Summary == "stopped" {
+			fmt.Println("Healstatus on items ===")
+			fmt.Println("Heal failed with", status.FailureDetail)
+			break
+		}
+
+		for _, item := range status.Items {
+			if err = enc.Encode(&item); err != nil {
+				log.Fatalln(err)
+			}
+		}
+
+		time.Sleep(time.Second)
+	}
+}

buildscripts/race.sh

@@ -3,6 +3,8 @@
 set -e
 
 export GORACE="history_size=7"
+export MINIO_API_REQUESTS_MAX=10000
+
 ## TODO remove `dsync` from race detector once this is merged and released https://go-review.googlesource.com/c/go/+/333529/
 for d in $(go list ./... | grep -v dsync); do
     CGO_ENABLED=1 go test -v -race --timeout 100m "$d"

buildscripts/resolve-right-versions.sh

@@ -39,7 +39,7 @@ function start_minio_5drive() {
     "${MINIO[@]}" --address ":$start_port" "${WORK_DIR}/cicd-corpus/disk{1...5}" > "${WORK_DIR}/server1.log" 2>&1 &
     pid=$!
     disown $pid
-    sleep 30
+    sleep 5
 
     if ! ps -p ${pid} 1>&2 >/dev/null; then
 	echo "server1 log:"

buildscripts/rewrite-old-new.sh

@@ -0,0 +1,151 @@
+#!/bin/bash -e
+
+set -E
+set -o pipefail
+set -x
+
+WORK_DIR="$PWD/.verify-$RANDOM"
+MINIO_CONFIG_DIR="$WORK_DIR/.minio"
+MINIO_OLD=( "$PWD/minio.RELEASE.2020-10-28T08-16-50Z" --config-dir "$MINIO_CONFIG_DIR" server )
+MINIO=( "$PWD/minio" --config-dir "$MINIO_CONFIG_DIR" server )
+
+if [ ! -x "$PWD/minio" ]; then
+    echo "minio executable binary not found in current directory"
+    exit 1
+fi
+
+function download_old_release() {
+    if [ ! -f minio.RELEASE.2020-10-28T08-16-50Z ]; then
+	curl --silent -O https://dl.minio.io/server/minio/release/linux-amd64/archive/minio.RELEASE.2020-10-28T08-16-50Z
+	chmod a+x minio.RELEASE.2020-10-28T08-16-50Z
+    fi
+}
+
+function verify_rewrite() {
+    start_port=$1
+
+    export MINIO_ACCESS_KEY=minio
+    export MINIO_SECRET_KEY=minio123
+    export MC_HOST_minio="http://minio:minio123@127.0.0.1:${start_port}/"
+    unset MINIO_KMS_AUTO_ENCRYPTION # do not auto-encrypt objects
+    export MINIO_CI_CD=1
+
+    MC_BUILD_DIR="mc-$RANDOM"
+    if ! git clone --quiet https://github.com/minio/mc "$MC_BUILD_DIR"; then
+	echo "failed to download https://github.com/minio/mc"
+	purge "${MC_BUILD_DIR}"
+	exit 1
+    fi
+
+    (cd "${MC_BUILD_DIR}" && go build -o "$WORK_DIR/mc")
+
+    # remove mc source.
+    purge "${MC_BUILD_DIR}"
+
+    "${MINIO_OLD[@]}" --address ":$start_port" "${WORK_DIR}/xl{1...16}" > "${WORK_DIR}/server1.log" 2>&1 &
+    pid=$!
+    disown $pid
+    sleep 10
+
+    if ! ps -p ${pid} 1>&2 >/dev/null; then
+	echo "server1 log:"
+	cat "${WORK_DIR}/server1.log"
+	echo "FAILED"
+	purge "$WORK_DIR"
+	exit 1
+    fi
+
+    "${WORK_DIR}/mc" mb minio/healing-rewrite-bucket --quiet --with-lock
+    "${WORK_DIR}/mc" cp \
+		     buildscripts/verify-build.sh \
+		     minio/healing-rewrite-bucket/ \
+		     --disable-multipart --quiet
+
+    "${WORK_DIR}/mc" cp \
+		     buildscripts/verify-build.sh \
+		     minio/healing-rewrite-bucket/ \
+		     --disable-multipart --quiet
+
+    "${WORK_DIR}/mc" cp \
+		     buildscripts/verify-build.sh \
+		     minio/healing-rewrite-bucket/ \
+		     --disable-multipart --quiet
+
+    kill ${pid}
+    sleep 3
+
+    "${MINIO[@]}" --address ":$start_port" "${WORK_DIR}/xl{1...16}" > "${WORK_DIR}/server1.log" 2>&1 &
+    pid=$!
+    disown $pid
+    sleep 10
+
+    if ! ps -p ${pid} 1>&2 >/dev/null; then
+	echo "server1 log:"
+	cat "${WORK_DIR}/server1.log"
+	echo "FAILED"
+	purge "$WORK_DIR"
+	exit 1
+    fi
+
+    go build ./docs/debugging/s3-check-md5/
+    if ! ./s3-check-md5 \
+	 -debug \
+	 -versions \
+	 -access-key minio \
+	 -secret-key minio123 \
+	 -endpoint http://127.0.0.1:${start_port}/ 2>&1 | grep INTACT; then
+	echo "server1 log:"
+	cat "${WORK_DIR}/server1.log"
+	echo "FAILED"
+	mkdir -p inspects
+	(cd inspects; "${WORK_DIR}/mc" admin inspect minio/healing-rewrite-bucket/verify-build.sh/**)
+
+	"${WORK_DIR}/mc" mb play/inspects
+	"${WORK_DIR}/mc" mirror inspects play/inspects
+
+	purge "$WORK_DIR"
+	exit 1
+    fi
+
+    go run ./buildscripts/heal-manual.go "127.0.0.1:${start_port}" "minio" "minio123"
+    sleep 1
+
+    if ! ./s3-check-md5 \
+	 -debug \
+	 -versions \
+	 -access-key minio \
+	 -secret-key minio123 \
+	 -endpoint http://127.0.0.1:${start_port}/ 2>&1 | grep INTACT; then
+	echo "server1 log:"
+	cat "${WORK_DIR}/server1.log"
+	echo "FAILED"
+	mkdir -p inspects
+	(cd inspects; "${WORK_DIR}/mc" admin inspect minio/healing-rewrite-bucket/verify-build.sh/**)
+
+	"${WORK_DIR}/mc" mb play/inspects
+	"${WORK_DIR}/mc" mirror inspects play/inspects
+
+	purge "$WORK_DIR"
+	exit 1
+    fi
+
+    kill ${pid}
+}
+
+function main() {
+    download_old_release
+
+    start_port=$(shuf -i 10000-65000 -n 1)
+
+    verify_rewrite ${start_port}
+}
+
+function purge()
+{
+    rm -rf "$1"
+}
+
+( main "$@" )
+rv=$?
+purge "$WORK_DIR"
+exit "$rv"

buildscripts/verify-healing-with-root-disks.sh

@@ -0,0 +1,97 @@
+#!/bin/bash -e
+
+set -E
+set -o pipefail
+set -x
+
+
+if [ ! -x "$PWD/minio" ]; then
+    echo "minio executable binary not found in current directory"
+    exit 1
+fi
+
+WORK_DIR="$(mktemp -d)"
+MINIO_CONFIG_DIR="$WORK_DIR/.minio"
+MINIO=( "$PWD/minio" --config-dir "$MINIO_CONFIG_DIR" server )
+
+
+function start_minio() {
+    start_port=$1
+
+    export MINIO_ROOT_USER=minio
+    export MINIO_ROOT_PASSWORD=minio123
+    unset MINIO_KMS_AUTO_ENCRYPTION # do not auto-encrypt objects
+    unset MINIO_CI_CD
+    unset CI
+
+    args=()
+    for i in $(seq 1 4); do
+            args+=("http://localhost:$[${start_port}+$i]${WORK_DIR}/mnt/disk$i/ ")
+    done
+
+    for i in $(seq 1 4); do
+            "${MINIO[@]}" --address ":$[$start_port+$i]" ${args[@]} 2>&1 >"${WORK_DIR}/server$i.log" &
+    done
+
+    # Wait until all nodes return 403
+    for i in $(seq 1 4); do
+	    while [ "$(curl -m 1 -s -o /dev/null -w "%{http_code}" http://localhost:$[$start_port+$i])" -ne "403" ]; do
+		    echo -n ".";
+                    sleep 1;
+            done
+    done
+
+ }
+
+# Prepare fake disks with losetup
+function prepare_block_devices() {
+        mkdir -p ${WORK_DIR}/disks/ ${WORK_DIR}/mnt/
+        for i in 1 2 3 4; do
+                dd if=/dev/zero of=${WORK_DIR}/disks/img.$i bs=1M count=2048
+                mkfs.ext4 -F ${WORK_DIR}/disks/img.$i
+                sudo mknod /dev/minio-loopdisk$i b 7 $[256-$i]
+                sudo losetup /dev/minio-loopdisk$i ${WORK_DIR}/disks/img.$i
+                mkdir -p ${WORK_DIR}/mnt/disk$i/
+                sudo mount /dev/minio-loopdisk$i ${WORK_DIR}/mnt/disk$i/
+                sudo chown "$(id -u):$(id -g)" /dev/minio-loopdisk$i ${WORK_DIR}/mnt/disk$i/
+        done
+}
+
+# Start a distributed MinIO setup, unmount one disk and check if it is formatted
+function main() {
+    start_port=$(shuf -i 10000-65000 -n 1)
+    start_minio ${start_port}
+
+    # Unmount the disk, after the unmount the device id
+    # /tmp/xxx/mnt/disk4 will be the same as '/' and it
+    # will be detected as root disk
+    while [ "$u" != "0" ]; do
+	    sudo umount ${WORK_DIR}/mnt/disk4/
+	    u=$?
+	    sleep 1
+    done
+
+    # Wait until MinIO self heal kicks in
+    sleep 60
+
+    if [ -f ${WORK_DIR}/mnt/disk4/.minio.sys/format.json ]; then
+	    echo "A root disk is formatted unexpectedely"
+            cat "${WORK_DIR}/server4.log"
+            exit -1
+    fi
+}
+
+function cleanup() {
+    pkill minio
+    sudo umount ${WORK_DIR}/mnt/disk{1..3}/
+    sudo rm /dev/minio-loopdisk*
+    rm -rf "$WORK_DIR"
+}
+
+( prepare_block_devices )
+( main "$@" )
+rv=$?
+
+cleanup
+exit "$rv"
+

cmd/acl-handlers.go

@@ -80,7 +80,7 @@ func (api objectAPIHandlers) PutBucketACLHandler(w http.ResponseWriter, r *http.
 	}
 
 	// Before proceeding validate if bucket exists.
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -142,7 +142,7 @@ func (api objectAPIHandlers) GetBucketACLHandler(w http.ResponseWriter, r *http.
 	}
 
 	// Before proceeding validate if bucket exists.
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return

cmd/admin-bucket-handlers.go

@@ -18,15 +18,31 @@
 package cmd
 
 import (
+	"bytes"
+	"encoding/base64"
 	"encoding/json"
+	"encoding/xml"
+	"errors"
+	"fmt"
 	"io"
 	"io/ioutil"
 	"net/http"
+	"strings"
+	"time"
 
 	"github.com/gorilla/mux"
 	jsoniter "github.com/json-iterator/go"
+	"github.com/klauspost/compress/zip"
+	"github.com/minio/kes"
 	"github.com/minio/madmin-go"
+	"github.com/minio/minio-go/v7/pkg/tags"
+	"github.com/minio/minio/internal/bucket/lifecycle"
+	objectlock "github.com/minio/minio/internal/bucket/object/lock"
+	"github.com/minio/minio/internal/bucket/versioning"
+	"github.com/minio/minio/internal/event"
+	"github.com/minio/minio/internal/kms"
 	"github.com/minio/minio/internal/logger"
+	"github.com/minio/pkg/bucket/policy"
 	iampolicy "github.com/minio/pkg/iam/policy"
 )
 
@@ -47,14 +63,13 @@ func (a adminAPIHandlers) PutBucketQuotaConfigHandler(w http.ResponseWriter, r *
 
 	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.SetBucketQuotaAdminAction)
 	if objectAPI == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
 
 	vars := mux.Vars(r)
 	bucket := pathClean(vars["bucket"])
 
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -76,15 +91,17 @@ func (a adminAPIHandlers) PutBucketQuotaConfigHandler(w http.ResponseWriter, r *
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketQuotaConfigFile, data); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketQuotaConfigFile, data)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
 	bucketMeta := madmin.SRBucketMeta{
-		Type:   madmin.SRBucketMetaTypeQuotaConfig,
-		Bucket: bucket,
-		Quota:  data,
+		Type:      madmin.SRBucketMetaTypeQuotaConfig,
+		Bucket:    bucket,
+		Quota:     data,
+		UpdatedAt: updatedAt,
 	}
 	if quotaConfig.Quota == 0 {
 		bucketMeta.Quota = nil
@@ -108,14 +125,13 @@ func (a adminAPIHandlers) GetBucketQuotaConfigHandler(w http.ResponseWriter, r *
 
 	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.GetBucketQuotaAdminAction)
 	if objectAPI == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
 
 	vars := mux.Vars(r)
 	bucket := pathClean(vars["bucket"])
 
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -153,12 +169,11 @@ func (a adminAPIHandlers) SetRemoteTargetHandler(w http.ResponseWriter, r *http.
 	// Get current object layer instance.
 	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.SetBucketTargetAction)
 	if objectAPI == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -234,7 +249,7 @@ func (a adminAPIHandlers) SetRemoteTargetHandler(w http.ResponseWriter, r *http.
 	}
 	if err = globalBucketTargetSys.SetTarget(ctx, bucket, &target, update); err != nil {
 		switch err.(type) {
-		case BucketRemoteConnectionErr:
+		case RemoteTargetConnectionErr:
 			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrReplicationRemoteConnectionError, err), r.URL)
 		default:
 			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
@@ -251,7 +266,7 @@ func (a adminAPIHandlers) SetRemoteTargetHandler(w http.ResponseWriter, r *http.
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err), r.URL)
 		return
 	}
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketTargetsFile, tgtBytes); err != nil {
+	if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketTargetsFile, tgtBytes); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -282,12 +297,11 @@ func (a adminAPIHandlers) ListRemoteTargetsHandler(w http.ResponseWriter, r *htt
 	// Get current object layer instance.
 	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.GetBucketTargetAction)
 	if objectAPI == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
 	if bucket != "" {
 		// Check if bucket exists.
-		if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+		if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 			return
 		}
@@ -322,12 +336,11 @@ func (a adminAPIHandlers) RemoveRemoteTargetHandler(w http.ResponseWriter, r *ht
 	// Get current object layer instance.
 	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.SetBucketTargetAction)
 	if objectAPI == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -346,7 +359,7 @@ func (a adminAPIHandlers) RemoveRemoteTargetHandler(w http.ResponseWriter, r *ht
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err), r.URL)
 		return
 	}
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketTargetsFile, tgtBytes); err != nil {
+	if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketTargetsFile, tgtBytes); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -354,3 +367,794 @@ func (a adminAPIHandlers) RemoveRemoteTargetHandler(w http.ResponseWriter, r *ht
 	// Write success response.
 	writeSuccessNoContent(w)
 }
+
+// ExportBucketMetadataHandler - exports all bucket metadata as a zipped file
+func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ExportBucketMetadata")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	bucket := pathClean(r.Form.Get("bucket"))
+	if globalIsGateway {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+	// Get current object layer instance.
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.ExportBucketMetadataAction)
+	if objectAPI == nil {
+		return
+	}
+
+	var (
+		buckets []BucketInfo
+		err     error
+	)
+	if bucket != "" {
+		// Check if bucket exists.
+		if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
+			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
+			return
+		}
+		buckets = append(buckets, BucketInfo{Name: bucket})
+	} else {
+		buckets, err = objectAPI.ListBuckets(ctx, BucketOptions{})
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+	}
+
+	// Initialize a zip writer which will provide a zipped content
+	// of bucket metadata
+	zipWriter := zip.NewWriter(w)
+	defer zipWriter.Close()
+	rawDataFn := func(r io.Reader, filename string, sz int) error {
+		header, zerr := zip.FileInfoHeader(dummyFileInfo{
+			name:    filename,
+			size:    int64(sz),
+			mode:    0o600,
+			modTime: time.Now(),
+			isDir:   false,
+			sys:     nil,
+		})
+		if zerr != nil {
+			logger.LogIf(ctx, zerr)
+			return nil
+		}
+		header.Method = zip.Deflate
+		zwriter, zerr := zipWriter.CreateHeader(header)
+		if zerr != nil {
+			logger.LogIf(ctx, zerr)
+			return nil
+		}
+		if _, err := io.Copy(zwriter, r); err != nil {
+			logger.LogIf(ctx, err)
+		}
+		return nil
+	}
+
+	cfgFiles := []string{
+		bucketPolicyConfig,
+		bucketNotificationConfig,
+		bucketLifecycleConfig,
+		bucketSSEConfig,
+		bucketTaggingConfig,
+		bucketQuotaConfigFile,
+		objectLockConfig,
+		bucketVersioningConfig,
+		bucketReplicationConfig,
+		bucketTargetsFile,
+	}
+	for _, bi := range buckets {
+		for _, cfgFile := range cfgFiles {
+			cfgPath := pathJoin(bi.Name, cfgFile)
+			bucket := bi.Name
+			switch cfgFile {
+			case bucketNotificationConfig:
+				config, err := globalBucketMetadataSys.GetNotificationConfig(bucket)
+				if err != nil {
+					logger.LogIf(ctx, err)
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketLifecycleConfig:
+				config, err := globalBucketMetadataSys.GetLifecycleConfig(bucket)
+				if err != nil {
+					if errors.Is(err, BucketLifecycleNotFound{Bucket: bucket}) {
+						continue
+					}
+					logger.LogIf(ctx, err)
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketQuotaConfigFile:
+				config, _, err := globalBucketMetadataSys.GetQuotaConfig(ctx, bucket)
+				if err != nil {
+					if errors.Is(err, BucketQuotaConfigNotFound{Bucket: bucket}) {
+						continue
+					}
+					writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+					return
+				}
+				configData, err := json.Marshal(config)
+				if err != nil {
+					writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketSSEConfig:
+				config, _, err := globalBucketMetadataSys.GetSSEConfig(bucket)
+				if err != nil {
+					if errors.Is(err, BucketSSEConfigNotFound{Bucket: bucket}) {
+						continue
+					}
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketTaggingConfig:
+				config, _, err := globalBucketMetadataSys.GetTaggingConfig(bucket)
+				if err != nil {
+					if errors.Is(err, BucketTaggingNotFound{Bucket: bucket}) {
+						continue
+					}
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case objectLockConfig:
+				config, _, err := globalBucketMetadataSys.GetObjectLockConfig(bucket)
+				if err != nil {
+					if errors.Is(err, BucketObjectLockConfigNotFound{Bucket: bucket}) {
+						continue
+					}
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketVersioningConfig:
+				config, _, err := globalBucketMetadataSys.GetVersioningConfig(bucket)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				// ignore empty versioning configs
+				if config.Status != versioning.Enabled && config.Status != versioning.Suspended {
+					continue
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketReplicationConfig:
+				config, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
+				if err != nil {
+					if errors.Is(err, BucketReplicationConfigNotFound{Bucket: bucket}) {
+						continue
+					}
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			case bucketTargetsFile:
+				config, err := globalBucketMetadataSys.GetBucketTargetsConfig(bucket)
+				if err != nil {
+					if errors.Is(err, BucketRemoteTargetNotFound{Bucket: bucket}) {
+						continue
+					}
+
+					writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+					return
+				}
+				configData, err := xml.Marshal(config)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
+					return
+				}
+			}
+		}
+	}
+}
+
+type importMetaReport struct {
+	madmin.BucketMetaImportErrs
+}
+
+func (i *importMetaReport) SetStatus(bucket, fname string, err error) {
+	st := i.Buckets[bucket]
+	var errMsg string
+	if err != nil {
+		errMsg = err.Error()
+	}
+	switch fname {
+	case bucketPolicyConfig:
+		st.Policy = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketNotificationConfig:
+		st.Notification = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketLifecycleConfig:
+		st.Lifecycle = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketSSEConfig:
+		st.SSEConfig = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketTaggingConfig:
+		st.Tagging = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketQuotaConfigFile:
+		st.Quota = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case objectLockConfig:
+		st.ObjectLock = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	case bucketVersioningConfig:
+		st.Versioning = madmin.MetaStatus{IsSet: true, Err: errMsg}
+	default:
+		st.Err = errMsg
+	}
+	i.Buckets[bucket] = st
+}
+
+// ImportBucketMetadataHandler - imports all bucket metadata from a zipped file and overwrite bucket metadata config
+// There are some caveats regarding the following:
+// 1. object lock config - object lock should have been specified at time of bucket creation. Only default retention settings are imported here.
+// 2. Replication config - is omitted from import as remote target credentials are not available from exported data for security reasons.
+// 3. lifecycle config - if transition rules are present, tier name needs to have been defined.
+func (a adminAPIHandlers) ImportBucketMetadataHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ImportBucketMetadata")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	if globalIsGateway {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+	// Get current object layer instance.
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.ImportBucketMetadataAction)
+	if objectAPI == nil {
+		return
+	}
+	data, err := ioutil.ReadAll(r.Body)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
+		return
+	}
+	reader := bytes.NewReader(data)
+	zr, err := zip.NewReader(reader, int64(len(data)))
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
+		return
+	}
+	bucketMap := make(map[string]struct{}, 1)
+	rpt := importMetaReport{
+		madmin.BucketMetaImportErrs{
+			Buckets: make(map[string]madmin.BucketStatus, len(zr.File)),
+		},
+	}
+	// import object lock config if any - order of import matters here.
+	for _, file := range zr.File {
+		slc := strings.Split(file.Name, slashSeparator)
+		if len(slc) != 2 { // expecting bucket/configfile in the zipfile
+			rpt.SetStatus(file.Name, "", fmt.Errorf("malformed zip - expecting format bucket/<config.json>"))
+			continue
+		}
+		bucket, fileName := slc[0], slc[1]
+		switch fileName {
+		case objectLockConfig:
+			reader, err := file.Open()
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			config, err := objectlock.ParseObjectLockConfig(reader)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s (%s)", errorCodes[ErrMalformedXML].Description, err))
+				continue
+			}
+
+			configData, err := xml.Marshal(config)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			if _, ok := bucketMap[bucket]; !ok {
+				opts := MakeBucketOptions{
+					LockEnabled: config.ObjectLockEnabled == "Enabled",
+				}
+				err = objectAPI.MakeBucketWithLocation(ctx, bucket, opts)
+				if err != nil {
+					if _, ok := err.(BucketExists); !ok {
+						rpt.SetStatus(bucket, fileName, err)
+						continue
+					}
+				}
+				bucketMap[bucket] = struct{}{}
+			}
+
+			// Deny object locking configuration settings on existing buckets without object lock enabled.
+			if _, _, err = globalBucketMetadataSys.GetObjectLockConfig(bucket); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, objectLockConfig, configData)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+
+			// Call site replication hook.
+			//
+			// We encode the xml bytes as base64 to ensure there are no encoding
+			// errors.
+			cfgStr := base64.StdEncoding.EncodeToString(configData)
+			if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
+				Type:             madmin.SRBucketMetaTypeObjectLockConfig,
+				Bucket:           bucket,
+				ObjectLockConfig: &cfgStr,
+				UpdatedAt:        updatedAt,
+			}); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+		}
+	}
+
+	// import versioning metadata
+	for _, file := range zr.File {
+		slc := strings.Split(file.Name, slashSeparator)
+		if len(slc) != 2 { // expecting bucket/configfile in the zipfile
+			rpt.SetStatus(file.Name, "", fmt.Errorf("malformed zip - expecting format bucket/<config.json>"))
+			continue
+		}
+		bucket, fileName := slc[0], slc[1]
+		switch fileName {
+		case bucketVersioningConfig:
+			reader, err := file.Open()
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			v, err := versioning.ParseConfig(io.LimitReader(reader, maxBucketVersioningConfigSize))
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			if _, ok := bucketMap[bucket]; !ok {
+				if err = objectAPI.MakeBucketWithLocation(ctx, bucket, MakeBucketOptions{}); err != nil {
+					if _, ok := err.(BucketExists); !ok {
+						rpt.SetStatus(bucket, fileName, err)
+						continue
+					}
+				}
+				bucketMap[bucket] = struct{}{}
+			}
+
+			if globalSiteReplicationSys.isEnabled() && v.Suspended() {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("Cluster replication is enabled for this site, so the versioning state cannot be suspended."))
+				continue
+			}
+
+			if rcfg, _ := globalBucketObjectLockSys.Get(bucket); rcfg.LockEnabled && v.Suspended() {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("An Object Lock configuration is present on this bucket, so the versioning state cannot be suspended."))
+				continue
+			}
+			if _, err := getReplicationConfig(ctx, bucket); err == nil && v.Suspended() {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("A replication configuration is present on this bucket, so the versioning state cannot be suspended."))
+				continue
+			}
+
+			configData, err := xml.Marshal(v)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s (%s)", errorCodes[ErrMalformedXML].Description, err))
+				continue
+			}
+
+			if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketVersioningConfig, configData); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+		}
+	}
+
+	for _, file := range zr.File {
+		reader, err := file.Open()
+		if err != nil {
+			rpt.SetStatus(file.Name, "", err)
+			continue
+		}
+		sz := file.FileInfo().Size()
+		slc := strings.Split(file.Name, slashSeparator)
+		if len(slc) != 2 { // expecting bucket/configfile in the zipfile
+			rpt.SetStatus(file.Name, "", fmt.Errorf("malformed zip - expecting format bucket/<config.json>"))
+			continue
+		}
+		bucket, fileName := slc[0], slc[1]
+		// create bucket if it does not exist yet.
+		if _, ok := bucketMap[bucket]; !ok {
+			err = objectAPI.MakeBucketWithLocation(ctx, bucket, MakeBucketOptions{})
+			if err != nil {
+				if _, ok := err.(BucketExists); !ok {
+					rpt.SetStatus(bucket, "", err)
+					continue
+				}
+			}
+			bucketMap[bucket] = struct{}{}
+		}
+		if _, ok := bucketMap[bucket]; !ok {
+			continue
+		}
+		switch fileName {
+		case bucketNotificationConfig:
+			config, err := event.ParseConfig(io.LimitReader(reader, sz), globalSite.Region, globalNotificationSys.targetList)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s (%s)", errorCodes[ErrMalformedXML].Description, err))
+				continue
+			}
+
+			configData, err := xml.Marshal(config)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketNotificationConfig, configData); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rulesMap := config.ToRulesMap()
+			globalNotificationSys.AddRulesMap(bucket, rulesMap)
+			rpt.SetStatus(bucket, fileName, nil)
+		case bucketPolicyConfig:
+			// Error out if Content-Length is beyond allowed size.
+			if sz > maxBucketPolicySize {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf(ErrPolicyTooLarge.String()))
+				continue
+			}
+
+			bucketPolicyBytes, err := ioutil.ReadAll(io.LimitReader(reader, sz))
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			bucketPolicy, err := policy.ParseConfig(bytes.NewReader(bucketPolicyBytes), bucket)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			// Version in policy must not be empty
+			if bucketPolicy.Version == "" {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf(ErrMalformedPolicy.String()))
+				continue
+			}
+
+			configData, err := json.Marshal(bucketPolicy)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketPolicyConfig, configData)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+			// Call site replication hook.
+			if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
+				Type:      madmin.SRBucketMetaTypePolicy,
+				Bucket:    bucket,
+				Policy:    bucketPolicyBytes,
+				UpdatedAt: updatedAt,
+			}); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+		case bucketLifecycleConfig:
+			bucketLifecycle, err := lifecycle.ParseLifecycleConfig(io.LimitReader(reader, sz))
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			// Validate the received bucket policy document
+			if err = bucketLifecycle.Validate(); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			// Validate the transition storage ARNs
+			if err = validateTransitionTier(bucketLifecycle); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			configData, err := xml.Marshal(bucketLifecycle)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketLifecycleConfig, configData); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+		case bucketSSEConfig:
+			// Parse bucket encryption xml
+			encConfig, err := validateBucketSSEConfig(io.LimitReader(reader, maxBucketSSEConfigSize))
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s (%s)", errorCodes[ErrMalformedXML].Description, err))
+				continue
+			}
+
+			// Return error if KMS is not initialized
+			if GlobalKMS == nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s", errorCodes[ErrKMSNotConfigured].Description))
+				continue
+			}
+			kmsKey := encConfig.KeyID()
+			if kmsKey != "" {
+				kmsContext := kms.Context{"MinIO admin API": "ServerInfoHandler"} // Context for a test key operation
+				_, err := GlobalKMS.GenerateKey(ctx, kmsKey, kmsContext)
+				if err != nil {
+					if errors.Is(err, kes.ErrKeyNotFound) {
+						rpt.SetStatus(bucket, fileName, errKMSKeyNotFound)
+						continue
+					}
+					rpt.SetStatus(bucket, fileName, err)
+					continue
+				}
+			}
+
+			configData, err := xml.Marshal(encConfig)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			// Store the bucket encryption configuration in the object layer
+			updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketSSEConfig, configData)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+
+			// Call site replication hook.
+			//
+			// We encode the xml bytes as base64 to ensure there are no encoding
+			// errors.
+			cfgStr := base64.StdEncoding.EncodeToString(configData)
+			if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
+				Type:      madmin.SRBucketMetaTypeSSEConfig,
+				Bucket:    bucket,
+				SSEConfig: &cfgStr,
+				UpdatedAt: updatedAt,
+			}); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+		case bucketTaggingConfig:
+			tags, err := tags.ParseBucketXML(io.LimitReader(reader, sz))
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("%s (%s)", errorCodes[ErrMalformedXML].Description, err))
+				continue
+			}
+
+			configData, err := xml.Marshal(tags)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketTaggingConfig, configData)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+
+			// Call site replication hook.
+			//
+			// We encode the xml bytes as base64 to ensure there are no encoding
+			// errors.
+			cfgStr := base64.StdEncoding.EncodeToString(configData)
+			if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
+				Type:      madmin.SRBucketMetaTypeTags,
+				Bucket:    bucket,
+				Tags:      &cfgStr,
+				UpdatedAt: updatedAt,
+			}); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+		case bucketQuotaConfigFile:
+			data, err := ioutil.ReadAll(reader)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			quotaConfig, err := parseBucketQuota(bucket, data)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+
+			if quotaConfig.Type == "fifo" {
+				rpt.SetStatus(bucket, fileName, fmt.Errorf("Detected older 'fifo' quota config, 'fifo' feature is removed and not supported anymore"))
+				continue
+			}
+
+			updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketQuotaConfigFile, data)
+			if err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+			rpt.SetStatus(bucket, fileName, nil)
+
+			bucketMeta := madmin.SRBucketMeta{
+				Type:      madmin.SRBucketMetaTypeQuotaConfig,
+				Bucket:    bucket,
+				Quota:     data,
+				UpdatedAt: updatedAt,
+			}
+			if quotaConfig.Quota == 0 {
+				bucketMeta.Quota = nil
+			}
+
+			// Call site replication hook.
+			if err = globalSiteReplicationSys.BucketMetaHook(ctx, bucketMeta); err != nil {
+				rpt.SetStatus(bucket, fileName, err)
+				continue
+			}
+		}
+	}
+
+	rptData, err := json.Marshal(rpt.BucketMetaImportErrs)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, rptData)
+}
+
+// ReplicationDiffHandler - POST returns info on unreplicated versions for a remote target ARN
+// to the connected HTTP client. This is a MinIO only extension
+func (a adminAPIHandlers) ReplicationDiffHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ReplicationDiff")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+
+	if globalIsGateway {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	// Get current object layer instance.
+	objectAPI := newObjectLayerFn()
+	if objectAPI == nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+	// check if user has permissions to perform this operation
+	if s3Error := checkRequestAuthType(ctx, r, policy.ListBucketVersionsAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	opts := extractReplicateDiffOpts(r.Form)
+	if opts.ARN != "" {
+		tgt := globalBucketTargetSys.GetRemoteBucketTargetByArn(ctx, bucket, opts.ARN)
+		if tgt.Empty() {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrInvalidRequest, fmt.Errorf("invalid arn : '%s'", opts.ARN)), r.URL)
+			return
+		}
+	}
+
+	keepAliveTicker := time.NewTicker(500 * time.Millisecond)
+	defer keepAliveTicker.Stop()
+
+	diffCh, err := getReplicationDiff(ctx, objectAPI, bucket, opts)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	enc := json.NewEncoder(w)
+	for {
+		select {
+		case entry, ok := <-diffCh:
+			if !ok {
+				return
+			}
+			if err := enc.Encode(entry); err != nil {
+				return
+			}
+			if len(diffCh) == 0 {
+				// Flush if nothing is queued
+				w.(http.Flusher).Flush()
+			}
+		case <-keepAliveTicker.C:
+			if len(diffCh) > 0 {
+				continue
+			}
+			if _, err := w.Write([]byte(" ")); err != nil {
+				return
+			}
+			w.(http.Flusher).Flush()
+		case <-ctx.Done():
+			return
+		}
+	}
+}

cmd/admin-handler-utils.go

@@ -20,6 +20,7 @@ package cmd
 import (
 	"context"
 	"errors"
+	"fmt"
 	"net/http"
 
 	"github.com/minio/kes"
@@ -29,6 +30,10 @@ import (
 	iampolicy "github.com/minio/pkg/iam/policy"
 )
 
+// validateAdminReq will validate request against and return whether it is allowed.
+// If any of the supplied actions are allowed it will be successful.
+// If nil ObjectLayer is returned, the operation is not permitted.
+// When nil ObjectLayer has been returned an error has always been sent to w.
 func validateAdminReq(ctx context.Context, w http.ResponseWriter, r *http.Request, actions ...iampolicy.AdminAction) (ObjectLayer, auth.Credentials) {
 	// Get current object layer instance.
 	objectAPI := newObjectLayerFn()
@@ -40,11 +45,16 @@ func validateAdminReq(ctx context.Context, w http.ResponseWriter, r *http.Reques
 	for _, action := range actions {
 		// Validate request signature.
 		cred, adminAPIErr := checkAdminRequestAuth(ctx, r, action, "")
-		if adminAPIErr != ErrNone {
+		switch adminAPIErr {
+		case ErrNone:
+			return objectAPI, cred
+		case ErrAccessDenied:
+			// Try another
+			continue
+		default:
 			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(adminAPIErr), r.URL)
 			return nil, cred
 		}
-		return objectAPI, cred
 	}
 	writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
 	return nil, auth.Credentials{}
@@ -96,6 +106,12 @@ func toAdminAPIErr(ctx context.Context, err error) APIError {
 		}
 	default:
 		switch {
+		case errors.Is(err, errTooManyPolicies):
+			apiErr = APIError{
+				Code:           "XMinioAdminInvalidRequest",
+				Description:    err.Error(),
+				HTTPStatusCode: http.StatusBadRequest,
+			}
 		case errors.Is(err, errDecommissionAlreadyRunning):
 			apiErr = APIError{
 				Code:           "XMinioDecommissionNotAllowed",
@@ -217,3 +233,27 @@ func toAdminAPIErrCode(ctx context.Context, err error) APIErrorCode {
 		return toAPIErrorCode(ctx, err)
 	}
 }
+
+// wraps export error for more context
+func exportError(ctx context.Context, err error, fname, entity string) APIError {
+	if entity == "" {
+		return toAPIError(ctx, fmt.Errorf("error exporting %s with: %w", fname, err))
+	}
+	return toAPIError(ctx, fmt.Errorf("error exporting %s from %s with: %w", entity, fname, err))
+}
+
+// wraps import error for more context
+func importError(ctx context.Context, err error, fname, entity string) APIError {
+	if entity == "" {
+		return toAPIError(ctx, fmt.Errorf("error importing %s with: %w", fname, err))
+	}
+	return toAPIError(ctx, fmt.Errorf("error importing %s from %s with: %w", entity, fname, err))
+}
+
+// wraps import error for more context
+func importErrorWithAPIErr(ctx context.Context, apiErr APIErrorCode, err error, fname, entity string) APIError {
+	if entity == "" {
+		return errorCodes.ToAPIErrWithErr(apiErr, fmt.Errorf("error importing %s with: %w", fname, err))
+	}
+	return errorCodes.ToAPIErrWithErr(apiErr, fmt.Errorf("error importing %s from %s with: %w", entity, fname, err))
+}

cmd/admin-handlers-config-kv.go

@@ -189,15 +189,20 @@ func (a adminAPIHandlers) GetConfigKVHandler(w http.ResponseWriter, r *http.Requ
 
 	cfg := globalServerConfig.Clone()
 	vars := mux.Vars(r)
-	buf := &bytes.Buffer{}
-	cw := config.NewConfigWriteTo(cfg, vars["key"])
-	if _, err := cw.WriteTo(buf); err != nil {
+	subSys := vars["key"]
+	subSysConfigs, err := cfg.GetSubsysInfo(subSys)
+	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
 
+	var s strings.Builder
+	for _, subSysConfig := range subSysConfigs {
+		subSysConfig.AddString(&s, false)
+	}
+
 	password := cred.SecretKey
-	econfigData, err := madmin.EncryptData(password, buf.Bytes())
+	econfigData, err := madmin.EncryptData(password, []byte(s.String()))
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -427,42 +432,30 @@ func (a adminAPIHandlers) GetConfigHandler(w http.ResponseWriter, r *http.Reques
 		count += len(cfg[hkv.Key])
 	}
 	for _, hkv := range hkvs {
-		v := cfg[hkv.Key]
-		for target, kv := range v {
-			off := kv.Get(config.Enable) == config.EnableOff
+		// We ignore the error below, as we cannot get one.
+		cfgSubsysItems, _ := cfg.GetSubsysInfo(hkv.Key)
+
+		for _, item := range cfgSubsysItems {
+			off := item.Config.Get(config.Enable) == config.EnableOff
 			switch hkv.Key {
 			case config.EtcdSubSys:
-				off = !etcd.Enabled(kv)
+				off = !etcd.Enabled(item.Config)
 			case config.CacheSubSys:
-				off = !cache.Enabled(kv)
+				off = !cache.Enabled(item.Config)
 			case config.StorageClassSubSys:
-				off = !storageclass.Enabled(kv)
+				off = !storageclass.Enabled(item.Config)
 			case config.PolicyPluginSubSys:
-				off = !polplugin.Enabled(kv)
+				off = !polplugin.Enabled(item.Config)
 			case config.IdentityOpenIDSubSys:
-				off = !openid.Enabled(kv)
+				off = !openid.Enabled(item.Config)
 			case config.IdentityLDAPSubSys:
-				off = !xldap.Enabled(kv)
+				off = !xldap.Enabled(item.Config)
 			case config.IdentityTLSSubSys:
 				off = !globalSTSTLSConfig.Enabled
 			case config.IdentityPluginSubSys:
-				off = !idplugin.Enabled(kv)
-			}
-			if off {
-				s.WriteString(config.KvComment)
-				s.WriteString(config.KvSpaceSeparator)
-			}
-			s.WriteString(hkv.Key)
-			if target != config.Default {
-				s.WriteString(config.SubSystemSeparator)
-				s.WriteString(target)
-			}
-			s.WriteString(config.KvSpaceSeparator)
-			s.WriteString(kv.String())
-			count--
-			if count > 0 {
-				s.WriteString(config.KvNewline)
+				off = !idplugin.Enabled(item.Config)
 			}
+			item.AddString(&s, off)
 		}
 	}
 

cmd/admin-handlers-idp-config.go

@@ -0,0 +1,321 @@
+// Copyright (c) 2015-2022 MinIO, Inc.
+//
+// This file is part of MinIO Object Storage stack
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+package cmd
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+
+	"github.com/gorilla/mux"
+	"github.com/minio/madmin-go"
+	"github.com/minio/minio-go/v7/pkg/set"
+	"github.com/minio/minio/internal/config"
+	"github.com/minio/minio/internal/config/identity/openid"
+	"github.com/minio/minio/internal/logger"
+	iampolicy "github.com/minio/pkg/iam/policy"
+)
+
+// List of implemented ID config types.
+var idCfgTypes = set.CreateStringSet("openid")
+
+// SetIdentityProviderCfg:
+//
+// PUT <admin-prefix>/id-cfg?type=openid&name=dex1
+func (a adminAPIHandlers) SetIdentityProviderCfg(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "SetIdentityCfg")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	objectAPI, cred := validateAdminReq(ctx, w, r, iampolicy.ConfigUpdateAdminAction)
+	if objectAPI == nil {
+		return
+	}
+
+	if r.ContentLength > maxEConfigJSONSize || r.ContentLength == -1 {
+		// More than maxConfigSize bytes were available
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigTooLarge), r.URL)
+		return
+	}
+
+	password := cred.SecretKey
+	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
+	if err != nil {
+		logger.LogIf(ctx, err, logger.Application)
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
+		return
+	}
+
+	cfgType := mux.Vars(r)["type"]
+	if !idCfgTypes.Contains(cfgType) {
+		// TODO: change this to invalid type error when implementation
+		// is complete.
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	var cfgDataBuilder strings.Builder
+	switch cfgType {
+	case "openid":
+		fmt.Fprintf(&cfgDataBuilder, "identity_openid")
+	}
+
+	// Ensure body content type is opaque.
+	contentType := r.Header.Get("Content-Type")
+	if contentType != "application/octet-stream" {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL)
+		return
+	}
+
+	// Subsystem configuration name could be empty.
+	cfgName := mux.Vars(r)["name"]
+	if cfgName != "" {
+		fmt.Fprintf(&cfgDataBuilder, "%s%s", config.SubSystemSeparator, cfgName)
+	}
+
+	fmt.Fprintf(&cfgDataBuilder, "%s%s", config.KvSpaceSeparator, string(reqBytes))
+
+	cfgData := cfgDataBuilder.String()
+	subSys, _, _, err := config.GetSubSys(cfgData)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	cfg, err := readServerConfig(ctx, objectAPI)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	dynamic, err := cfg.ReadConfig(strings.NewReader(cfgData))
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	// IDP config is not dynamic. Sanity check.
+	if dynamic {
+		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInternalError), err.Error(), r.URL)
+		return
+	}
+
+	if err = validateConfig(cfg, subSys); err != nil {
+		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), err.Error(), r.URL)
+		return
+	}
+
+	// Update the actual server config on disk.
+	if err = saveServerConfig(ctx, objectAPI, cfg); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	// Write to the config input KV to history.
+	if err = saveServerConfigHistory(ctx, objectAPI, []byte(cfgData)); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseHeadersOnly(w)
+}
+
+// GetIdentityProviderCfg:
+//
+// GET <admin-prefix>/id-cfg?type=openid&name=dex_test
+//
+// GetIdentityProviderCfg returns a list of configured IDPs on the server if
+// name is empty. If name is non-empty, returns the configuration details for
+// the IDP of the given type and configuration name. The configuration name for
+// the default ("un-named") configuration target is `_`.
+func (a adminAPIHandlers) GetIdentityProviderCfg(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "GetIdentityProviderCfg")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	objectAPI, cred := validateAdminReq(ctx, w, r, iampolicy.ConfigUpdateAdminAction)
+	if objectAPI == nil {
+		return
+	}
+
+	cfgType := mux.Vars(r)["type"]
+	cfgName := r.Form.Get("name")
+	password := cred.SecretKey
+
+	if !idCfgTypes.Contains(cfgType) {
+		// TODO: change this to invalid type error when implementation
+		// is complete.
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	// If no cfgName is provided, we list.
+	if cfgName == "" {
+		a.listIdentityProviders(ctx, w, r, cfgType, password)
+		return
+	}
+
+	cfg := globalServerConfig.Clone()
+
+	cfgInfos, err := globalOpenIDConfig.GetConfigInfo(cfg, cfgName)
+	if err != nil {
+		if err == openid.ErrProviderConfigNotFound {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminNoSuchConfigTarget), r.URL)
+			return
+		}
+
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	res := madmin.IDPConfig{
+		Type: cfgType,
+		Name: cfgName,
+		Info: cfgInfos,
+	}
+	data, err := json.Marshal(res)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	econfigData, err := madmin.EncryptData(password, data)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, econfigData)
+}
+
+func (a adminAPIHandlers) listIdentityProviders(ctx context.Context, w http.ResponseWriter, r *http.Request, cfgType, password string) {
+	// var subSys string
+	switch cfgType {
+	case "openid":
+		// subSys = config.IdentityOpenIDSubSys
+	default:
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	cfg := globalServerConfig.Clone()
+	cfgList, err := globalOpenIDConfig.GetConfigList(cfg)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	data, err := json.Marshal(cfgList)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	econfigData, err := madmin.EncryptData(password, data)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, econfigData)
+}
+
+// DeleteIdentityProviderCfg:
+//
+// DELETE <admin-prefix>/id-cfg?type=openid&name=dex_test
+func (a adminAPIHandlers) DeleteIdentityProviderCfg(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "DeleteIdentityProviderCfg")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.ConfigUpdateAdminAction)
+	if objectAPI == nil {
+		return
+	}
+
+	cfgType := mux.Vars(r)["type"]
+	cfgName := mux.Vars(r)["name"]
+	if !idCfgTypes.Contains(cfgType) {
+		// TODO: change this to invalid type error when implementation
+		// is complete.
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	cfg := globalServerConfig.Clone()
+
+	cfgInfos, err := globalOpenIDConfig.GetConfigInfo(cfg, cfgName)
+	if err != nil {
+		if err == openid.ErrProviderConfigNotFound {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminNoSuchConfigTarget), r.URL)
+			return
+		}
+
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	hasEnv := false
+	for _, ci := range cfgInfos {
+		if ci.IsCfg && ci.IsEnv {
+			hasEnv = true
+			break
+		}
+	}
+
+	if hasEnv {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigEnvOverridden), r.URL)
+		return
+	}
+
+	var subSys string
+	switch cfgType {
+	case "openid":
+		subSys = config.IdentityOpenIDSubSys
+	default:
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	cfg, err = readServerConfig(ctx, objectAPI)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	if err = cfg.DelKVS(fmt.Sprintf("%s:%s", subSys, cfgName)); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+	if err = validateConfig(cfg, subSys); err != nil {
+		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), err.Error(), r.URL)
+		return
+	}
+	if err = saveServerConfig(ctx, objectAPI, cfg); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	dynamic := config.SubSystemsDynamic.Contains(subSys)
+	if dynamic {
+		applyDynamic(ctx, objectAPI, cfg, subSys, r, w)
+	}
+}

cmd/admin-handlers-pools.go

@@ -19,6 +19,7 @@ package cmd
 
 import (
 	"encoding/json"
+	"fmt"
 	"net/http"
 
 	"github.com/gorilla/mux"
@@ -149,8 +150,10 @@ func (a adminAPIHandlers) StatusPool(w http.ResponseWriter, r *http.Request) {
 
 	idx := globalEndpoints.GetPoolIdx(v)
 	if idx == -1 {
+		apiErr := toAdminAPIErr(ctx, errInvalidArgument)
+		apiErr.Description = fmt.Sprintf("specified pool '%s' not found, please specify a valid pool", v)
 		// We didn't find any matching pools, invalid input
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errInvalidArgument), r.URL)
+		writeErrorResponseJSON(ctx, w, apiErr, r.URL)
 		return
 	}
 

cmd/admin-handlers-site-replication.go

@@ -24,6 +24,8 @@ import (
 	"io"
 	"io/ioutil"
 	"net/http"
+	"strings"
+	"time"
 
 	"github.com/gorilla/mux"
 	"github.com/minio/madmin-go"
@@ -116,19 +118,38 @@ func (a adminAPIHandlers) SRPeerBucketOps(w http.ResponseWriter, r *http.Request
 		_, isLockEnabled := r.Form["lockEnabled"]
 		_, isVersioningEnabled := r.Form["versioningEnabled"]
 		_, isForceCreate := r.Form["forceCreate"]
-		opts := BucketOptions{
+		createdAtStr := strings.TrimSpace(r.Form.Get("createdAt"))
+		createdAt, cerr := time.Parse(time.RFC3339Nano, createdAtStr)
+		if cerr != nil {
+			createdAt = timeSentinel
+		}
+
+		opts := MakeBucketOptions{
 			Location:          r.Form.Get("location"),
 			LockEnabled:       isLockEnabled,
 			VersioningEnabled: isVersioningEnabled,
 			ForceCreate:       isForceCreate,
+			CreatedAt:         createdAt,
 		}
 		err = globalSiteReplicationSys.PeerBucketMakeWithVersioningHandler(ctx, bucket, opts)
 	case madmin.ConfigureReplBktOp:
 		err = globalSiteReplicationSys.PeerBucketConfigureReplHandler(ctx, bucket)
 	case madmin.DeleteBucketBktOp:
-		err = globalSiteReplicationSys.PeerBucketDeleteHandler(ctx, bucket, false)
+		_, noRecreate := r.Form["noRecreate"]
+		err = globalSiteReplicationSys.PeerBucketDeleteHandler(ctx, bucket, DeleteBucketOptions{
+			Force:      false,
+			NoRecreate: noRecreate,
+			SRDeleteOp: getSRBucketDeleteOp(true),
+		})
 	case madmin.ForceDeleteBucketBktOp:
-		err = globalSiteReplicationSys.PeerBucketDeleteHandler(ctx, bucket, true)
+		_, noRecreate := r.Form["noRecreate"]
+		err = globalSiteReplicationSys.PeerBucketDeleteHandler(ctx, bucket, DeleteBucketOptions{
+			Force:      true,
+			NoRecreate: noRecreate,
+			SRDeleteOp: getSRBucketDeleteOp(true),
+		})
+	case madmin.PurgeDeletedBucketOp:
+		globalSiteReplicationSys.purgeDeletedBucket(ctx, objectAPI, bucket)
 	}
 	if err != nil {
 		logger.LogIf(ctx, err)
@@ -160,7 +181,7 @@ func (a adminAPIHandlers) SRPeerReplicateIAMItem(w http.ResponseWriter, r *http.
 		err = errSRInvalidRequest(errInvalidArgument)
 	case madmin.SRIAMItemPolicy:
 		if item.Policy == nil {
-			err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, nil)
+			err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, nil, item.UpdatedAt)
 		} else {
 			policy, perr := iampolicy.ParseConfig(bytes.NewReader(item.Policy))
 			if perr != nil {
@@ -168,21 +189,21 @@ func (a adminAPIHandlers) SRPeerReplicateIAMItem(w http.ResponseWriter, r *http.
 				return
 			}
 			if policy.IsEmpty() {
-				err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, nil)
+				err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, nil, item.UpdatedAt)
 			} else {
-				err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, policy)
+				err = globalSiteReplicationSys.PeerAddPolicyHandler(ctx, item.Name, policy, item.UpdatedAt)
 			}
 		}
 	case madmin.SRIAMItemSvcAcc:
-		err = globalSiteReplicationSys.PeerSvcAccChangeHandler(ctx, item.SvcAccChange)
+		err = globalSiteReplicationSys.PeerSvcAccChangeHandler(ctx, item.SvcAccChange, item.UpdatedAt)
 	case madmin.SRIAMItemPolicyMapping:
-		err = globalSiteReplicationSys.PeerPolicyMappingHandler(ctx, item.PolicyMapping)
+		err = globalSiteReplicationSys.PeerPolicyMappingHandler(ctx, item.PolicyMapping, item.UpdatedAt)
 	case madmin.SRIAMItemSTSAcc:
-		err = globalSiteReplicationSys.PeerSTSAccHandler(ctx, item.STSCredential)
+		err = globalSiteReplicationSys.PeerSTSAccHandler(ctx, item.STSCredential, item.UpdatedAt)
 	case madmin.SRIAMItemIAMUser:
-		err = globalSiteReplicationSys.PeerIAMUserChangeHandler(ctx, item.IAMUser)
+		err = globalSiteReplicationSys.PeerIAMUserChangeHandler(ctx, item.IAMUser, item.UpdatedAt)
 	case madmin.SRIAMItemGroupInfo:
-		err = globalSiteReplicationSys.PeerGroupInfoChangeHandler(ctx, item.GroupInfo)
+		err = globalSiteReplicationSys.PeerGroupInfoChangeHandler(ctx, item.GroupInfo, item.UpdatedAt)
 	}
 	if err != nil {
 		logger.LogIf(ctx, err)
@@ -214,7 +235,7 @@ func (a adminAPIHandlers) SRPeerReplicateBucketItem(w http.ResponseWriter, r *ht
 		err = errSRInvalidRequest(errInvalidArgument)
 	case madmin.SRBucketMetaTypePolicy:
 		if item.Policy == nil {
-			err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, nil)
+			err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, nil, item.UpdatedAt)
 		} else {
 			bktPolicy, berr := policy.ParseConfig(bytes.NewReader(item.Policy), item.Bucket)
 			if berr != nil {
@@ -222,33 +243,33 @@ func (a adminAPIHandlers) SRPeerReplicateBucketItem(w http.ResponseWriter, r *ht
 				return
 			}
 			if bktPolicy.IsEmpty() {
-				err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, nil)
+				err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, nil, item.UpdatedAt)
 			} else {
-				err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, bktPolicy)
+				err = globalSiteReplicationSys.PeerBucketPolicyHandler(ctx, item.Bucket, bktPolicy, item.UpdatedAt)
 			}
 		}
 	case madmin.SRBucketMetaTypeQuotaConfig:
 		if item.Quota == nil {
-			err = globalSiteReplicationSys.PeerBucketQuotaConfigHandler(ctx, item.Bucket, nil)
+			err = globalSiteReplicationSys.PeerBucketQuotaConfigHandler(ctx, item.Bucket, nil, item.UpdatedAt)
 		} else {
 			quotaConfig, err := parseBucketQuota(item.Bucket, item.Quota)
 			if err != nil {
 				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 				return
 			}
-			if err = globalSiteReplicationSys.PeerBucketQuotaConfigHandler(ctx, item.Bucket, quotaConfig); err != nil {
+			if err = globalSiteReplicationSys.PeerBucketQuotaConfigHandler(ctx, item.Bucket, quotaConfig, item.UpdatedAt); err != nil {
 				writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 				return
 			}
 		}
 	case madmin.SRBucketMetaTypeVersionConfig:
-		err = globalSiteReplicationSys.PeerBucketVersioningHandler(ctx, item.Bucket, item.Versioning)
+		err = globalSiteReplicationSys.PeerBucketVersioningHandler(ctx, item.Bucket, item.Versioning, item.UpdatedAt)
 	case madmin.SRBucketMetaTypeTags:
-		err = globalSiteReplicationSys.PeerBucketTaggingHandler(ctx, item.Bucket, item.Tags)
+		err = globalSiteReplicationSys.PeerBucketTaggingHandler(ctx, item.Bucket, item.Tags, item.UpdatedAt)
 	case madmin.SRBucketMetaTypeObjectLockConfig:
-		err = globalSiteReplicationSys.PeerBucketObjectLockConfigHandler(ctx, item.Bucket, item.ObjectLockConfig)
+		err = globalSiteReplicationSys.PeerBucketObjectLockConfigHandler(ctx, item.Bucket, item.ObjectLockConfig, item.UpdatedAt)
 	case madmin.SRBucketMetaTypeSSEConfig:
-		err = globalSiteReplicationSys.PeerBucketSSEConfigHandler(ctx, item.Bucket, item.SSEConfig)
+		err = globalSiteReplicationSys.PeerBucketSSEConfigHandler(ctx, item.Bucket, item.SSEConfig, item.UpdatedAt)
 	}
 	if err != nil {
 		logger.LogIf(ctx, err)
@@ -436,6 +457,7 @@ func getSRStatusOptions(r *http.Request) (opts madmin.SRStatusOptions) {
 	opts.Users = q.Get("users") == "true"
 	opts.Entity = madmin.GetSREntityType(q.Get("entity"))
 	opts.EntityValue = q.Get("entityvalue")
+	opts.ShowDeleted = q.Get("showDeleted") == "true"
 	return
 }
 

cmd/admin-handlers-users.go

@@ -24,9 +24,12 @@ import (
 	"io"
 	"io/ioutil"
 	"net/http"
+	"os"
 	"sort"
+	"time"
 
 	"github.com/gorilla/mux"
+	"github.com/klauspost/compress/zip"
 	"github.com/minio/madmin-go"
 	"github.com/minio/minio/internal/auth"
 	"github.com/minio/minio/internal/config/dns"
@@ -69,6 +72,7 @@ func (a adminAPIHandlers) RemoveUser(w http.ResponseWriter, r *http.Request) {
 			AccessKey:   accessKey,
 			IsDeleteReq: true,
 		},
+		UpdatedAt: UTCNow(),
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -237,9 +241,9 @@ func (a adminAPIHandlers) UpdateGroupMembers(w http.ResponseWriter, r *http.Requ
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
 		return
 	}
-
+	var updatedAt time.Time
 	if updReq.IsRemove {
-		err = globalIAMSys.RemoveUsersFromGroup(ctx, updReq.Group, updReq.Members)
+		updatedAt, err = globalIAMSys.RemoveUsersFromGroup(ctx, updReq.Group, updReq.Members)
 	} else {
 		// Check if group already exists
 		if _, gerr := globalIAMSys.GetGroupDescription(updReq.Group); gerr != nil {
@@ -250,7 +254,7 @@ func (a adminAPIHandlers) UpdateGroupMembers(w http.ResponseWriter, r *http.Requ
 				return
 			}
 		}
-		err = globalIAMSys.AddUsersToGroup(ctx, updReq.Group, updReq.Members)
+		updatedAt, err = globalIAMSys.AddUsersToGroup(ctx, updReq.Group, updReq.Members)
 	}
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -262,6 +266,7 @@ func (a adminAPIHandlers) UpdateGroupMembers(w http.ResponseWriter, r *http.Requ
 		GroupInfo: &madmin.SRGroupInfo{
 			UpdateReq: updReq,
 		},
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -338,12 +343,15 @@ func (a adminAPIHandlers) SetGroupStatus(w http.ResponseWriter, r *http.Request)
 	group := vars["group"]
 	status := vars["status"]
 
-	var err error
+	var (
+		err       error
+		updatedAt time.Time
+	)
 	switch status {
 	case statusEnabled:
-		err = globalIAMSys.SetGroupStatus(ctx, group, true)
+		updatedAt, err = globalIAMSys.SetGroupStatus(ctx, group, true)
 	case statusDisabled:
-		err = globalIAMSys.SetGroupStatus(ctx, group, false)
+		updatedAt, err = globalIAMSys.SetGroupStatus(ctx, group, false)
 	default:
 		err = errInvalidArgument
 	}
@@ -361,6 +369,7 @@ func (a adminAPIHandlers) SetGroupStatus(w http.ResponseWriter, r *http.Request)
 				IsRemove: false,
 			},
 		},
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -388,7 +397,8 @@ func (a adminAPIHandlers) SetUserStatus(w http.ResponseWriter, r *http.Request)
 		return
 	}
 
-	if err := globalIAMSys.SetUserStatus(ctx, accessKey, madmin.AccountStatus(status)); err != nil {
+	updatedAt, err := globalIAMSys.SetUserStatus(ctx, accessKey, madmin.AccountStatus(status))
+	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
@@ -402,6 +412,7 @@ func (a adminAPIHandlers) SetUserStatus(w http.ResponseWriter, r *http.Request)
 				Status: madmin.AccountStatus(status),
 			},
 		},
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -436,8 +447,8 @@ func (a adminAPIHandlers) AddUser(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
-	userCred, exists := globalIAMSys.GetUser(ctx, accessKey)
-	if exists && (userCred.IsTemp() || userCred.IsServiceAccount()) {
+	user, exists := globalIAMSys.GetUser(ctx, accessKey)
+	if exists && (user.Credentials.IsTemp() || user.Credentials.IsServiceAccount()) {
 		// Updating STS credential is not allowed, and this API does not
 		// support updating service accounts.
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAddUserInvalidArgument), r.URL)
@@ -498,7 +509,8 @@ func (a adminAPIHandlers) AddUser(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
-	if err = globalIAMSys.CreateUser(ctx, accessKey, ureq); err != nil {
+	updatedAt, err := globalIAMSys.CreateUser(ctx, accessKey, ureq)
+	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
@@ -510,6 +522,7 @@ func (a adminAPIHandlers) AddUser(w http.ResponseWriter, r *http.Request) {
 			IsDeleteReq: false,
 			UserReq:     &ureq,
 		},
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -629,7 +642,7 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 			opts.claims[k] = v
 		}
 	} else {
-		// Need permission if we are creating a service acccount for a
+		// Need permission if we are creating a service account for a
 		// user <> to the request sender
 		if !globalIAMSys.IsAllowed(iampolicy.Args{
 			AccountName:     requestorUser,
@@ -671,7 +684,7 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 	}
 
 	opts.sessionPolicy = sp
-	newCred, err := globalIAMSys.NewServiceAccount(ctx, targetUser, targetGroups, opts)
+	newCred, updatedAt, err := globalIAMSys.NewServiceAccount(ctx, targetUser, targetGroups, opts)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -714,6 +727,7 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 					Status:        auth.AccountOn,
 				},
 			},
+			UpdatedAt: updatedAt,
 		})
 		if err != nil {
 			logger.LogIf(ctx, err)
@@ -797,7 +811,7 @@ func (a adminAPIHandlers) UpdateServiceAccount(w http.ResponseWriter, r *http.Re
 		status:        updateReq.NewStatus,
 		sessionPolicy: sp,
 	}
-	err = globalIAMSys.UpdateServiceAccount(ctx, accessKey, opts)
+	updatedAt, err := globalIAMSys.UpdateServiceAccount(ctx, accessKey, opts)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -815,6 +829,7 @@ func (a adminAPIHandlers) UpdateServiceAccount(w http.ResponseWriter, r *http.Re
 					SessionPolicy: updateReq.NewPolicy,
 				},
 			},
+			UpdatedAt: updatedAt,
 		})
 		if err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -1054,6 +1069,7 @@ func (a adminAPIHandlers) DeleteServiceAccount(w http.ResponseWriter, r *http.Re
 					AccessKey: serviceAccount,
 				},
 			},
+			UpdatedAt: UTCNow(),
 		}); err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
@@ -1161,7 +1177,7 @@ func (a adminAPIHandlers) AccountInfoHandler(w http.ResponseWriter, r *http.Requ
 			return buckets[i].Name < buckets[j].Name
 		})
 	} else {
-		buckets, err = objectAPI.ListBuckets(ctx)
+		buckets, err = objectAPI.ListBuckets(ctx, BucketOptions{})
 		if err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
@@ -1173,17 +1189,32 @@ func (a adminAPIHandlers) AccountInfoHandler(w http.ResponseWriter, r *http.Requ
 		// For derived credentials, check the parent user's permissions.
 		accountName = cred.ParentUser
 	}
-	policies, err := globalIAMSys.PolicyDBGet(accountName, false, cred.Groups...)
-	if err != nil {
-		logger.LogIf(ctx, err)
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
-	}
 
-	buf, err := json.MarshalIndent(globalIAMSys.GetCombinedPolicy(policies...), "", " ")
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
+	var buf []byte
+	if accountName == globalActiveCred.AccessKey {
+		for _, policy := range iampolicy.DefaultPolicies {
+			if policy.Name == "consoleAdmin" {
+				buf, err = json.MarshalIndent(policy.Definition, "", " ")
+				if err != nil {
+					writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+					return
+				}
+				break
+			}
+		}
+	} else {
+		policies, err := globalIAMSys.PolicyDBGet(accountName, false, cred.Groups...)
+		if err != nil {
+			logger.LogIf(ctx, err)
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+
+		buf, err = json.MarshalIndent(globalIAMSys.GetCombinedPolicy(policies...), "", " ")
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
 	}
 
 	acctInfo := madmin.AccountInfo{
@@ -1392,8 +1423,9 @@ func (a adminAPIHandlers) RemoveCannedPolicy(w http.ResponseWriter, r *http.Requ
 	// Call cluster-replication policy creation hook to replicate policy deletion to
 	// other minio clusters.
 	if err := globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
-		Type: madmin.SRIAMItemPolicy,
-		Name: policyName,
+		Type:      madmin.SRIAMItemPolicy,
+		Name:      policyName,
+		UpdatedAt: UTCNow(),
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -1450,7 +1482,8 @@ func (a adminAPIHandlers) AddCannedPolicy(w http.ResponseWriter, r *http.Request
 		return
 	}
 
-	if err = globalIAMSys.SetPolicy(ctx, policyName, *iamPolicy); err != nil {
+	updatedAt, err := globalIAMSys.SetPolicy(ctx, policyName, *iamPolicy)
+	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
@@ -1458,9 +1491,10 @@ func (a adminAPIHandlers) AddCannedPolicy(w http.ResponseWriter, r *http.Request
 	// Call cluster-replication policy creation hook to replicate policy to
 	// other minio clusters.
 	if err := globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
-		Type:   madmin.SRIAMItemPolicy,
-		Name:   policyName,
-		Policy: iamPolicyBytes,
+		Type:      madmin.SRIAMItemPolicy,
+		Name:      policyName,
+		Policy:    iamPolicyBytes,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -1495,7 +1529,8 @@ func (a adminAPIHandlers) SetPolicyForUserOrGroup(w http.ResponseWriter, r *http
 		}
 	}
 
-	if err := globalIAMSys.PolicyDBSet(ctx, entityName, policyName, isGroup); err != nil {
+	updatedAt, err := globalIAMSys.PolicyDBSet(ctx, entityName, policyName, isGroup)
+	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
@@ -1507,8 +1542,680 @@ func (a adminAPIHandlers) SetPolicyForUserOrGroup(w http.ResponseWriter, r *http
 			IsGroup:     isGroup,
 			Policy:      policyName,
 		},
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
 }
+
+const (
+	allPoliciesFile            = "policies.json"
+	allUsersFile               = "users.json"
+	allGroupsFile              = "groups.json"
+	allSvcAcctsFile            = "svcaccts.json"
+	userPolicyMappingsFile     = "user_mappings.json"
+	groupPolicyMappingsFile    = "group_mappings.json"
+	stsUserPolicyMappingsFile  = "stsuser_mappings.json"
+	stsGroupPolicyMappingsFile = "stsgroup_mappings.json"
+	iamAssetsDir               = "iam-assets"
+)
+
+// ExportIAMHandler - exports all iam info as a zipped file
+func (a adminAPIHandlers) ExportIAM(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ExportIAM")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	// Get current object layer instance.
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.ExportIAMAction)
+	if objectAPI == nil {
+		return
+	}
+	// Initialize a zip writer which will provide a zipped content
+	// of bucket metadata
+	zipWriter := zip.NewWriter(w)
+	defer zipWriter.Close()
+	rawDataFn := func(r io.Reader, filename string, sz int) error {
+		header, zerr := zip.FileInfoHeader(dummyFileInfo{
+			name:    filename,
+			size:    int64(sz),
+			mode:    0o600,
+			modTime: time.Now(),
+			isDir:   false,
+			sys:     nil,
+		})
+		if zerr != nil {
+			logger.LogIf(ctx, zerr)
+			return nil
+		}
+		header.Method = zip.Deflate
+		zwriter, zerr := zipWriter.CreateHeader(header)
+		if zerr != nil {
+			logger.LogIf(ctx, zerr)
+			return nil
+		}
+		if _, err := io.Copy(zwriter, r); err != nil {
+			logger.LogIf(ctx, err)
+		}
+		return nil
+	}
+
+	iamFiles := []string{
+		allPoliciesFile,
+		allUsersFile,
+		allGroupsFile,
+		allSvcAcctsFile,
+		userPolicyMappingsFile,
+		groupPolicyMappingsFile,
+		stsUserPolicyMappingsFile,
+		stsGroupPolicyMappingsFile,
+	}
+	for _, f := range iamFiles {
+		iamFile := pathJoin(iamAssetsDir, f)
+		switch f {
+		case allPoliciesFile:
+			allPolicies, err := globalIAMSys.ListPolicies(ctx, "")
+			if err != nil {
+				logger.LogIf(ctx, err)
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			policiesData, err := json.Marshal(allPolicies)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			if err = rawDataFn(bytes.NewReader(policiesData), iamFile, len(policiesData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case allUsersFile:
+			userIdentities := make(map[string]UserIdentity)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadUsers(ctx, regUser, userIdentities)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			userAccounts := make(map[string]madmin.AddOrUpdateUserReq)
+			for u, uid := range userIdentities {
+				status := madmin.AccountDisabled
+				if uid.Credentials.IsValid() {
+					status = madmin.AccountEnabled
+				}
+				userAccounts[u] = madmin.AddOrUpdateUserReq{
+					SecretKey: uid.Credentials.SecretKey,
+					Status:    status,
+				}
+			}
+			userData, err := json.Marshal(userAccounts)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			if err = rawDataFn(bytes.NewReader(userData), iamFile, len(userData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case allGroupsFile:
+			groups := make(map[string]GroupInfo)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadGroups(ctx, groups)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			groupData, err := json.Marshal(groups)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			if err = rawDataFn(bytes.NewReader(groupData), iamFile, len(groupData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case allSvcAcctsFile:
+			serviceAccounts := make(map[string]UserIdentity)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadUsers(ctx, svcUser, serviceAccounts)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			svcAccts := make(map[string]madmin.SRSvcAccCreate)
+			for user, acc := range serviceAccounts {
+				claims, err := globalIAMSys.GetClaimsForSvcAcc(ctx, acc.Credentials.AccessKey)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+					return
+				}
+				_, policy, err := globalIAMSys.GetServiceAccount(ctx, acc.Credentials.AccessKey)
+				if err != nil {
+					writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+					return
+				}
+
+				var policyJSON []byte
+				if policy != nil {
+					policyJSON, err = json.Marshal(policy)
+					if err != nil {
+						writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+						return
+					}
+				}
+				svcAccts[user] = madmin.SRSvcAccCreate{
+					Parent:        acc.Credentials.ParentUser,
+					AccessKey:     user,
+					SecretKey:     acc.Credentials.SecretKey,
+					Groups:        acc.Credentials.Groups,
+					Claims:        claims,
+					SessionPolicy: json.RawMessage(policyJSON),
+					Status:        acc.Credentials.Status,
+				}
+			}
+
+			svcAccData, err := json.Marshal(svcAccts)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			if err = rawDataFn(bytes.NewReader(svcAccData), iamFile, len(svcAccData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case userPolicyMappingsFile:
+			userPolicyMap := make(map[string]MappedPolicy)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadMappedPolicies(ctx, regUser, false, userPolicyMap)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			userPolData, err := json.Marshal(userPolicyMap)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			if err = rawDataFn(bytes.NewReader(userPolData), iamFile, len(userPolData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case groupPolicyMappingsFile:
+			groupPolicyMap := make(map[string]MappedPolicy)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadMappedPolicies(ctx, regUser, true, groupPolicyMap)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			grpPolData, err := json.Marshal(groupPolicyMap)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+
+			if err = rawDataFn(bytes.NewReader(grpPolData), iamFile, len(grpPolData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case stsUserPolicyMappingsFile:
+			userPolicyMap := make(map[string]MappedPolicy)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadMappedPolicies(ctx, stsUser, false, userPolicyMap)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			userPolData, err := json.Marshal(userPolicyMap)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			if err = rawDataFn(bytes.NewReader(userPolData), iamFile, len(userPolData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		case stsGroupPolicyMappingsFile:
+			groupPolicyMap := make(map[string]MappedPolicy)
+			globalIAMSys.store.rlock()
+			err := globalIAMSys.store.loadMappedPolicies(ctx, stsUser, true, groupPolicyMap)
+			globalIAMSys.store.runlock()
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			grpPolData, err := json.Marshal(groupPolicyMap)
+			if err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+			if err = rawDataFn(bytes.NewReader(grpPolData), iamFile, len(grpPolData)); err != nil {
+				writeErrorResponse(ctx, w, exportError(ctx, err, iamFile, ""), r.URL)
+				return
+			}
+		}
+	}
+}
+
+// ImportIAM - imports all IAM info into MinIO
+func (a adminAPIHandlers) ImportIAM(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ImportIAM")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	// Get current object layer instance.
+	objectAPI := newObjectLayerFn()
+	if objectAPI == nil || globalNotificationSys == nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+	cred, claims, owner, s3Err := validateAdminSignature(ctx, r, "")
+	if s3Err != ErrNone {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(s3Err), r.URL)
+		return
+	}
+	data, err := ioutil.ReadAll(r.Body)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
+		return
+	}
+	reader := bytes.NewReader(data)
+	zr, err := zip.NewReader(reader, int64(len(data)))
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
+		return
+	}
+	// import policies first
+	{
+
+		f, err := zr.Open(pathJoin(iamAssetsDir, allPoliciesFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allPoliciesFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var allPolicies map[string]iampolicy.Policy
+			data, err = ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allPoliciesFile, ""), r.URL)
+				return
+			}
+			err = json.Unmarshal(data, &allPolicies)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, allPoliciesFile, ""), r.URL)
+				return
+			}
+			for policyName, policy := range allPolicies {
+				if policy.IsEmpty() {
+					err = globalIAMSys.DeletePolicy(ctx, policyName, true)
+				} else {
+					_, err = globalIAMSys.SetPolicy(ctx, policyName, policy)
+				}
+				if err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, allPoliciesFile, policyName), r.URL)
+					return
+				}
+			}
+		}
+	}
+
+	// import users
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, allUsersFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allUsersFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var userAccts map[string]madmin.AddOrUpdateUserReq
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allUsersFile, ""), r.URL)
+				return
+			}
+			err = json.Unmarshal(data, &userAccts)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, allUsersFile, ""), r.URL)
+				return
+			}
+			for accessKey, ureq := range userAccts {
+				// Not allowed to add a user with same access key as root credential
+				if owner && accessKey == cred.AccessKey {
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAddUserInvalidArgument, err, allUsersFile, accessKey), r.URL)
+					return
+				}
+
+				user, exists := globalIAMSys.GetUser(ctx, accessKey)
+				if exists && (user.Credentials.IsTemp() || user.Credentials.IsServiceAccount()) {
+					// Updating STS credential is not allowed, and this API does not
+					// support updating service accounts.
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAddUserInvalidArgument, err, allUsersFile, accessKey), r.URL)
+					return
+				}
+
+				if (cred.IsTemp() || cred.IsServiceAccount()) && cred.ParentUser == accessKey {
+					// Incoming access key matches parent user then we should
+					// reject password change requests.
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAddUserInvalidArgument, err, allUsersFile, accessKey), r.URL)
+					return
+				}
+
+				// Check if accessKey has beginning and end space characters, this only applies to new users.
+				if !exists && hasSpaceBE(accessKey) {
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminResourceInvalidArgument, err, allUsersFile, accessKey), r.URL)
+					return
+				}
+
+				checkDenyOnly := false
+				if accessKey == cred.AccessKey {
+					// Check that there is no explicit deny - otherwise it's allowed
+					// to change one's own password.
+					checkDenyOnly = true
+				}
+
+				if !globalIAMSys.IsAllowed(iampolicy.Args{
+					AccountName:     cred.AccessKey,
+					Groups:          cred.Groups,
+					Action:          iampolicy.CreateUserAdminAction,
+					ConditionValues: getConditionValues(r, "", cred.AccessKey, claims),
+					IsOwner:         owner,
+					Claims:          claims,
+					DenyOnly:        checkDenyOnly,
+				}) {
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAccessDenied, err, allUsersFile, accessKey), r.URL)
+					return
+				}
+				if _, err = globalIAMSys.CreateUser(ctx, accessKey, ureq); err != nil {
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, toAdminAPIErrCode(ctx, err), err, allUsersFile, accessKey), r.URL)
+					return
+				}
+
+			}
+		}
+	}
+
+	// import groups
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, allGroupsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allGroupsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var grpInfos map[string]GroupInfo
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allGroupsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &grpInfos); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, allGroupsFile, ""), r.URL)
+				return
+			}
+			for group, grpInfo := range grpInfos {
+				// Check if group already exists
+				if _, gerr := globalIAMSys.GetGroupDescription(group); gerr != nil {
+					// If group does not exist, then check if the group has beginning and end space characters
+					// we will reject such group names.
+					if errors.Is(gerr, errNoSuchGroup) && hasSpaceBE(group) {
+						writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminResourceInvalidArgument, err, allGroupsFile, group), r.URL)
+						return
+					}
+				}
+				if _, gerr := globalIAMSys.AddUsersToGroup(ctx, group, grpInfo.Members); gerr != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, allGroupsFile, group), r.URL)
+					return
+				}
+			}
+		}
+	}
+
+	// import service accounts
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, allSvcAcctsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allSvcAcctsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var serviceAcctReqs map[string]madmin.SRSvcAccCreate
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, allSvcAcctsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &serviceAcctReqs); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, allSvcAcctsFile, ""), r.URL)
+				return
+			}
+			for user, svcAcctReq := range serviceAcctReqs {
+				var sp *iampolicy.Policy
+				var err error
+				if len(svcAcctReq.SessionPolicy) > 0 {
+					sp, err = iampolicy.ParseConfig(bytes.NewReader(svcAcctReq.SessionPolicy))
+					if err != nil {
+						writeErrorResponseJSON(ctx, w, importError(ctx, err, allSvcAcctsFile, user), r.URL)
+						return
+					}
+				}
+				// service account access key cannot have space characters beginning and end of the string.
+				if hasSpaceBE(svcAcctReq.AccessKey) {
+					writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+					return
+				}
+				if !globalIAMSys.IsAllowed(iampolicy.Args{
+					AccountName:     svcAcctReq.AccessKey,
+					Groups:          svcAcctReq.Groups,
+					Action:          iampolicy.CreateServiceAccountAdminAction,
+					ConditionValues: getConditionValues(r, "", cred.AccessKey, claims),
+					IsOwner:         owner,
+					Claims:          claims,
+				}) {
+					writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAccessDenied, err, allSvcAcctsFile, user), r.URL)
+					return
+				}
+				updateReq := true
+				_, _, err = globalIAMSys.GetServiceAccount(ctx, svcAcctReq.AccessKey)
+				if err != nil {
+					if !errors.Is(err, errNoSuchServiceAccount) {
+						writeErrorResponseJSON(ctx, w, importError(ctx, err, allSvcAcctsFile, user), r.URL)
+						return
+					}
+					updateReq = false
+				}
+				if updateReq {
+					opts := updateServiceAccountOpts{
+						secretKey:     svcAcctReq.SecretKey,
+						status:        svcAcctReq.Status,
+						sessionPolicy: sp,
+					}
+					_, err = globalIAMSys.UpdateServiceAccount(ctx, svcAcctReq.AccessKey, opts)
+					if err != nil {
+						writeErrorResponseJSON(ctx, w, importError(ctx, err, allSvcAcctsFile, user), r.URL)
+						return
+					}
+					continue
+				}
+				opts := newServiceAccountOpts{
+					accessKey:     user,
+					secretKey:     svcAcctReq.SecretKey,
+					sessionPolicy: sp,
+					claims:        svcAcctReq.Claims,
+				}
+
+				// In case of LDAP we need to resolve the targetUser to a DN and
+				// query their groups:
+				if globalLDAPConfig.Enabled {
+					opts.claims[ldapUserN] = svcAcctReq.AccessKey // simple username
+					targetUser, _, err := globalLDAPConfig.LookupUserDN(svcAcctReq.AccessKey)
+					if err != nil {
+						writeErrorResponseJSON(ctx, w, importError(ctx, err, allSvcAcctsFile, user), r.URL)
+						return
+					}
+					opts.claims[ldapUser] = targetUser // username DN
+				}
+
+				if _, _, err = globalIAMSys.NewServiceAccount(ctx, svcAcctReq.Parent, svcAcctReq.Groups, opts); err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, allSvcAcctsFile, user), r.URL)
+					return
+				}
+
+			}
+		}
+	}
+
+	// import user policy mappings
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, userPolicyMappingsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, userPolicyMappingsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var userPolicyMap map[string]MappedPolicy
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, userPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &userPolicyMap); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, userPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			for u, pm := range userPolicyMap {
+				// disallow setting policy mapping if user is a temporary user
+				ok, _, err := globalIAMSys.IsTempUser(u)
+				if err != nil && err != errNoSuchUser {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, userPolicyMappingsFile, u), r.URL)
+					return
+				}
+				if ok {
+					writeErrorResponseJSON(ctx, w, importError(ctx, errIAMActionNotAllowed, userPolicyMappingsFile, u), r.URL)
+					return
+				}
+				if _, err := globalIAMSys.PolicyDBSet(ctx, u, pm.Policies, false); err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, userPolicyMappingsFile, u), r.URL)
+					return
+				}
+			}
+		}
+	}
+
+	// import group policy mappings
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, groupPolicyMappingsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, groupPolicyMappingsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var grpPolicyMap map[string]MappedPolicy
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, groupPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &grpPolicyMap); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, groupPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			for g, pm := range grpPolicyMap {
+				if _, err := globalIAMSys.PolicyDBSet(ctx, g, pm.Policies, true); err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, groupPolicyMappingsFile, g), r.URL)
+					return
+				}
+			}
+		}
+	}
+
+	// import sts user policy mappings
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, stsUserPolicyMappingsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, stsUserPolicyMappingsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var userPolicyMap map[string]MappedPolicy
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, stsUserPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &userPolicyMap); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, stsUserPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			for u, pm := range userPolicyMap {
+				// disallow setting policy mapping if user is a temporary user
+				ok, _, err := globalIAMSys.IsTempUser(u)
+				if err != nil && err != errNoSuchUser {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, stsUserPolicyMappingsFile, u), r.URL)
+					return
+				}
+				if ok {
+					writeErrorResponseJSON(ctx, w, importError(ctx, errIAMActionNotAllowed, stsUserPolicyMappingsFile, u), r.URL)
+					return
+				}
+				if _, err := globalIAMSys.PolicyDBSet(ctx, u, pm.Policies, false); err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, stsUserPolicyMappingsFile, u), r.URL)
+					return
+				}
+			}
+		}
+	}
+
+	// import sts group policy mappings
+	{
+		f, err := zr.Open(pathJoin(iamAssetsDir, stsGroupPolicyMappingsFile))
+		switch {
+		case errors.Is(err, os.ErrNotExist):
+		case err != nil:
+			writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, stsGroupPolicyMappingsFile, ""), r.URL)
+			return
+		default:
+			defer f.Close()
+			var grpPolicyMap map[string]MappedPolicy
+			data, err := ioutil.ReadAll(f)
+			if err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrInvalidRequest, err, stsGroupPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			if err = json.Unmarshal(data, &grpPolicyMap); err != nil {
+				writeErrorResponseJSON(ctx, w, importErrorWithAPIErr(ctx, ErrAdminConfigBadJSON, err, stsGroupPolicyMappingsFile, ""), r.URL)
+				return
+			}
+			for g, pm := range grpPolicyMap {
+				if _, err := globalIAMSys.PolicyDBSet(ctx, g, pm.Policies, true); err != nil {
+					writeErrorResponseJSON(ctx, w, importError(ctx, err, stsGroupPolicyMappingsFile, g), r.URL)
+					return
+				}
+			}
+		}
+	}
+}

cmd/admin-handlers.go

@@ -28,6 +28,7 @@ import (
 	"hash/crc32"
 	"io"
 	"io/ioutil"
+	"math"
 	"math/rand"
 	"net/http"
 	"net/url"
@@ -38,7 +39,6 @@ import (
 	"sort"
 	"strconv"
 	"strings"
-	"sync"
 	"time"
 
 	"github.com/dustin/go-humanize"
@@ -51,6 +51,7 @@ import (
 	"github.com/minio/minio/internal/kms"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/minio/internal/logger/message/log"
+	"github.com/minio/minio/internal/pubsub"
 	iampolicy "github.com/minio/pkg/iam/policy"
 	xnet "github.com/minio/pkg/net"
 	"github.com/secure-io/sio-go"
@@ -70,16 +71,6 @@ const (
 	mgmtForceStop   = "forceStop"
 )
 
-func updateServer(u *url.URL, sha256Sum []byte, lrTime time.Time, releaseInfo string, mode string) (us madmin.ServerUpdateStatus, err error) {
-	if err = doUpdate(u, lrTime, sha256Sum, releaseInfo, mode); err != nil {
-		return us, err
-	}
-
-	us.CurrentVersion = Version
-	us.UpdatedVersion = lrTime.Format(minioReleaseTagTimeLayout)
-	return us, nil
-}
-
 // ServerUpdateHandler - POST /minio/admin/v3/update?updateURL={updateURL}
 // ----------
 // updates all minio servers and restarts them gracefully.
@@ -151,7 +142,16 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	for _, nerr := range globalNotificationSys.ServerUpdate(ctx, u, sha256Sum, lrTime, releaseInfo) {
+	// Download Binary Once
+	reader, err := downloadBinary(u, mode)
+	if err != nil {
+		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	// Push binary to other servers
+	for _, nerr := range globalNotificationSys.VerifyBinary(ctx, u, sha256Sum, releaseInfo, reader) {
 		if nerr.Err != nil {
 			err := AdminError{
 				Code:       AdminUpdateApplyFailure,
@@ -165,13 +165,39 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		}
 	}
 
-	updateStatus, err := updateServer(u, sha256Sum, lrTime, releaseInfo, mode)
+	err = verifyBinary(u, sha256Sum, releaseInfo, mode, reader)
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
 
+	for _, nerr := range globalNotificationSys.CommitBinary(ctx) {
+		if nerr.Err != nil {
+			err := AdminError{
+				Code:       AdminUpdateApplyFailure,
+				Message:    nerr.Err.Error(),
+				StatusCode: http.StatusInternalServerError,
+			}
+			logger.GetReqInfo(ctx).SetTags("peerAddress", nerr.Host.String())
+			logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+	}
+
+	err = commitBinary()
+	if err != nil {
+		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	updateStatus := madmin.ServerUpdateStatus{
+		CurrentVersion: Version,
+		UpdatedVersion: lrTime.Format(minioReleaseTagTimeLayout),
+	}
+
 	// Marshal API response
 	jsonBytes, err := json.Marshal(updateStatus)
 	if err != nil {
@@ -275,6 +301,8 @@ type ServerConnStats struct {
 	Throughput       uint64 `json:"throughput,omitempty"`
 	S3InputBytes     uint64 `json:"transferredS3"`
 	S3OutputBytes    uint64 `json:"receivedS3"`
+	AdminInputBytes  uint64 `json:"transferredAdmin"`
+	AdminOutputBytes uint64 `json:"receivedAdmin"`
 }
 
 // ServerHTTPAPIStats holds total number of HTTP operations from/to the server,
@@ -291,6 +319,8 @@ type ServerHTTPStats struct {
 	CurrentS3Requests      ServerHTTPAPIStats `json:"currentS3Requests"`
 	TotalS3Requests        ServerHTTPAPIStats `json:"totalS3Requests"`
 	TotalS3Errors          ServerHTTPAPIStats `json:"totalS3Errors"`
+	TotalS35xxErrors       ServerHTTPAPIStats `json:"totalS35xxErrors"`
+	TotalS34xxErrors       ServerHTTPAPIStats `json:"totalS34xxErrors"`
 	TotalS3Canceled        ServerHTTPAPIStats `json:"totalS3Canceled"`
 	TotalS3RejectedAuth    uint64             `json:"totalS3RejectedAuth"`
 	TotalS3RejectedTime    uint64             `json:"totalS3RejectedTime"`
@@ -340,6 +370,109 @@ func (a adminAPIHandlers) StorageInfoHandler(w http.ResponseWriter, r *http.Requ
 	writeSuccessResponseJSON(w, jsonBytes)
 }
 
+// MetricsHandler - GET /minio/admin/v3/metrics
+// ----------
+// Get realtime server metrics
+func (a adminAPIHandlers) MetricsHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "Metrics")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.ServerInfoAdminAction)
+	if objectAPI == nil {
+		return
+	}
+	const defaultMetricsInterval = time.Second
+
+	interval, err := time.ParseDuration(r.Form.Get("interval"))
+	if err != nil || interval < time.Second {
+		interval = defaultMetricsInterval
+	}
+	n, err := strconv.Atoi(r.Form.Get("n"))
+	if err != nil || n <= 0 {
+		n = math.MaxInt32
+	}
+	var types madmin.MetricType
+	if t, _ := strconv.ParseUint(r.Form.Get("types"), 10, 64); t != 0 {
+		types = madmin.MetricType(t)
+	} else {
+		types = madmin.MetricsAll
+	}
+
+	disks := strings.Split(r.Form.Get("disks"), ",")
+	byDisk := strings.EqualFold(r.Form.Get("by-disk"), "true")
+	var diskMap map[string]struct{}
+	if len(disks) > 0 && disks[0] != "" {
+		diskMap = make(map[string]struct{}, len(disks))
+		for _, k := range disks {
+			if k != "" {
+				diskMap[k] = struct{}{}
+			}
+		}
+	}
+
+	hosts := strings.Split(r.Form.Get("hosts"), ",")
+	byHost := strings.EqualFold(r.Form.Get("by-host"), "true")
+	var hostMap map[string]struct{}
+	if len(hosts) > 0 && hosts[0] != "" {
+		hostMap = make(map[string]struct{}, len(hosts))
+		for _, k := range hosts {
+			if k != "" {
+				hostMap[k] = struct{}{}
+			}
+		}
+	}
+
+	done := ctx.Done()
+	ticker := time.NewTicker(interval)
+	defer ticker.Stop()
+	w.Header().Set(xhttp.ContentType, string(mimeJSON))
+
+	for n > 0 {
+		var m madmin.RealtimeMetrics
+		mLocal := collectLocalMetrics(types, hostMap, diskMap)
+		m.Merge(&mLocal)
+
+		// Allow half the interval for collecting remote...
+		cctx, cancel := context.WithTimeout(ctx, interval/2)
+		mRemote := collectRemoteMetrics(cctx, types, hostMap, diskMap)
+		cancel()
+		m.Merge(&mRemote)
+		if !byHost {
+			m.ByHost = nil
+		}
+		if !byDisk {
+			m.ByDisk = nil
+		}
+
+		m.Final = n <= 1
+		// Marshal API response
+		jsonBytes, err := json.Marshal(m)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+		_, err = w.Write(jsonBytes)
+		if err != nil {
+			n = 0
+		}
+
+		n--
+		if n <= 0 {
+			break
+		}
+
+		// Flush before waiting for next...
+		w.(http.Flusher).Flush()
+
+		select {
+		case <-ticker.C:
+		case <-done:
+			return
+		}
+	}
+}
+
 // DataUsageInfoHandler - GET /minio/admin/v3/datausage
 // ----------
 // Get server/cluster data usage info
@@ -1058,17 +1191,17 @@ func (a adminAPIHandlers) NetperfHandler(w http.ResponseWriter, r *http.Request)
 	}
 }
 
-// SpeedtestHandler - Deprecated. See ObjectSpeedtestHandler
-func (a adminAPIHandlers) SpeedtestHandler(w http.ResponseWriter, r *http.Request) {
-	a.ObjectSpeedtestHandler(w, r)
+// SpeedtestHandler - Deprecated. See ObjectSpeedTestHandler
+func (a adminAPIHandlers) SpeedTestHandler(w http.ResponseWriter, r *http.Request) {
+	a.ObjectSpeedTestHandler(w, r)
 }
 
-// ObjectSpeedtestHandler - reports maximum speed of a cluster by performing PUT and
+// ObjectSpeedTestHandler - reports maximum speed of a cluster by performing PUT and
 // GET operations on the server, supports auto tuning by default by automatically
 // increasing concurrency and stopping when we have reached the limits on the
 // system.
-func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "ObjectSpeedtestHandler")
+func (a adminAPIHandlers) ObjectSpeedTestHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ObjectSpeedTestHandler")
 
 	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
 
@@ -1085,8 +1218,9 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 	sizeStr := r.Form.Get(peerRESTSize)
 	durationStr := r.Form.Get(peerRESTDuration)
 	concurrentStr := r.Form.Get(peerRESTConcurrent)
+	storageClass := strings.TrimSpace(r.Form.Get(peerRESTStorageClass))
+	customBucket := strings.TrimSpace(r.Form.Get(peerRESTBucket))
 	autotune := r.Form.Get("autotune") == "true"
-	storageClass := r.Form.Get("storage-class")
 
 	size, err := strconv.Atoi(sizeStr)
 	if err != nil {
@@ -1098,17 +1232,14 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 		concurrent = 32
 	}
 
-	if runtime.GOMAXPROCS(0) < concurrent {
-		concurrent = runtime.GOMAXPROCS(0)
-	}
-
 	duration, err := time.ParseDuration(durationStr)
 	if err != nil {
 		duration = time.Second * 10
 	}
 
-	sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, objectAPI, concurrent, size, autotune)
+	storageInfo, _ := objectAPI.StorageInfo(ctx)
 
+	sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, storageInfo, concurrent, size, autotune)
 	if !sufficientCapacity {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, AdminError{
 			Code:       "XMinioSpeedtestInsufficientCapacity",
@@ -1122,15 +1253,23 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 		autotune = false
 	}
 
-	bucketExists, err := makeObjectPerfBucket(ctx, objectAPI)
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
-		return
+	if customBucket == "" {
+		customBucket = globalObjectPerfBucket
+
+		bucketExists, err := makeObjectPerfBucket(ctx, objectAPI, customBucket)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
+			return
+		}
+
+		if !bucketExists {
+			defer deleteObjectPerfBucket(objectAPI)
+		}
 	}
 
-	if !bucketExists {
-		defer deleteObjectPerfBucket(objectAPI)
-	}
+	defer objectAPI.DeleteObject(ctx, customBucket, speedTest+SlashSeparator, ObjectOptions{
+		DeletePrefix: true,
+	})
 
 	// Freeze all incoming S3 API calls before running speedtest.
 	globalNotificationSys.ServiceFreeze(ctx, true)
@@ -1142,15 +1281,32 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 	defer keepAliveTicker.Stop()
 
 	enc := json.NewEncoder(w)
-	ch := objectSpeedTest(ctx, speedTestOpts{size, concurrent, duration, autotune, storageClass})
+	ch := objectSpeedTest(ctx, speedTestOpts{
+		objectSize:       size,
+		concurrencyStart: concurrent,
+		duration:         duration,
+		autotune:         autotune,
+		storageClass:     storageClass,
+		bucketName:       customBucket,
+	})
+	var prevResult madmin.SpeedTestResult
 	for {
 		select {
 		case <-ctx.Done():
 			return
 		case <-keepAliveTicker.C:
-			// Write a blank entry to prevent client from disconnecting
-			if err := enc.Encode(madmin.SpeedTestResult{}); err != nil {
-				return
+			// if previous result is set keep writing the
+			// previous result back to the client
+			if prevResult.Version != "" {
+				if err := enc.Encode(prevResult); err != nil {
+					return
+				}
+			} else {
+				// first result is not yet obtained, keep writing
+				// empty entry to prevent client from disconnecting.
+				if err := enc.Encode(madmin.SpeedTestResult{}); err != nil {
+					return
+				}
 			}
 			w.(http.Flusher).Flush()
 		case result, ok := <-ch:
@@ -1160,14 +1316,14 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 			if err := enc.Encode(result); err != nil {
 				return
 			}
+			prevResult = result
 			w.(http.Flusher).Flush()
 		}
 	}
 }
 
-func makeObjectPerfBucket(ctx context.Context, objectAPI ObjectLayer) (bucketExists bool, err error) {
-	err = objectAPI.MakeBucketWithLocation(ctx, globalObjectPerfBucket, BucketOptions{})
-	if err != nil {
+func makeObjectPerfBucket(ctx context.Context, objectAPI ObjectLayer, bucketName string) (bucketExists bool, err error) {
+	if err = objectAPI.MakeBucketWithLocation(ctx, bucketName, MakeBucketOptions{}); err != nil {
 		if _, ok := err.(BucketExists); !ok {
 			// Only BucketExists error can be ignored.
 			return false, err
@@ -1181,13 +1337,13 @@ func deleteObjectPerfBucket(objectAPI ObjectLayer) {
 	objectAPI.DeleteBucket(context.Background(), globalObjectPerfBucket, DeleteBucketOptions{
 		Force:      true,
 		NoRecreate: true,
+		SRDeleteOp: getSRBucketDeleteOp(globalSiteReplicationSys.isEnabled()),
 	})
 }
 
-func validateObjPerfOptions(ctx context.Context, objectAPI ObjectLayer, concurrent int, size int, autotune bool) (sufficientCapacity bool, canAutotune bool, capacityErrMsg string) {
-	storageInfo, _ := objectAPI.StorageInfo(ctx)
+func validateObjPerfOptions(ctx context.Context, storageInfo madmin.StorageInfo, concurrent int, size int, autotune bool) (bool, bool, string) {
 	capacityNeeded := uint64(concurrent * size)
-	capacity := uint64(GetTotalUsableCapacityFree(storageInfo.Disks, storageInfo))
+	capacity := GetTotalUsableCapacityFree(storageInfo.Disks, storageInfo)
 
 	if capacity < capacityNeeded {
 		return false, false, fmt.Sprintf("not enough usable space available to perform speedtest - expected %s, got %s",
@@ -1261,43 +1417,29 @@ func (a adminAPIHandlers) DriveSpeedtestHandler(w http.ResponseWriter, r *http.R
 	keepAliveTicker := time.NewTicker(500 * time.Millisecond)
 	defer keepAliveTicker.Stop()
 
-	enc := json.NewEncoder(w)
 	ch := globalNotificationSys.DriveSpeedTest(ctx, opts)
 
-	var wg sync.WaitGroup
-	wg.Add(1)
-
-	// local driveSpeedTest
-	go func() {
-		defer wg.Done()
-		enc.Encode(driveSpeedTest(ctx, opts))
-		if wf, ok := w.(http.Flusher); ok {
-			wf.Flush()
-		}
-	}()
-
+	enc := json.NewEncoder(w)
 	for {
 		select {
 		case <-ctx.Done():
-			goto endloop
+			return
 		case <-keepAliveTicker.C:
 			// Write a blank entry to prevent client from disconnecting
 			if err := enc.Encode(madmin.DriveSpeedTestResult{}); err != nil {
-				goto endloop
+				return
 			}
 			w.(http.Flusher).Flush()
 		case result, ok := <-ch:
 			if !ok {
-				goto endloop
+				return
 			}
 			if err := enc.Encode(result); err != nil {
-				goto endloop
+				return
 			}
 			w.(http.Flusher).Flush()
 		}
 	}
-endloop:
-	wg.Wait()
 }
 
 // Admin API errors
@@ -1312,72 +1454,45 @@ const (
 // - input entry is not of the type *madmin.TraceInfo*
 // - errOnly entries are to be traced, not status code 2xx, 3xx.
 // - madmin.TraceInfo type is asked by opts
-func mustTrace(entry interface{}, opts madmin.ServiceTraceOpts) (shouldTrace bool) {
-	trcInfo, ok := entry.(madmin.TraceInfo)
-	if !ok {
+func shouldTrace(trcInfo madmin.TraceInfo, opts madmin.ServiceTraceOpts) (shouldTrace bool) {
+	// Reject all unwanted types.
+	want := opts.TraceTypes()
+	if !want.Contains(trcInfo.TraceType) {
 		return false
 	}
 
-	// Override shouldTrace decision with errOnly filtering
-	defer func() {
-		if shouldTrace && opts.OnlyErrors {
-			shouldTrace = trcInfo.RespInfo.StatusCode >= http.StatusBadRequest
-		}
-	}()
+	isHTTP := trcInfo.TraceType.Overlaps(madmin.TraceInternal|madmin.TraceS3) && trcInfo.HTTP != nil
 
-	if opts.Threshold > 0 {
-		var latency time.Duration
-		switch trcInfo.TraceType {
-		case madmin.TraceOS:
-			latency = trcInfo.OSStats.Duration
-		case madmin.TraceStorage:
-			latency = trcInfo.StorageStats.Duration
-		case madmin.TraceHTTP:
-			latency = trcInfo.CallStats.Latency
-		}
-		if latency < opts.Threshold {
-			return false
-		}
+	// Check latency...
+	if opts.Threshold > 0 && trcInfo.Duration < opts.Threshold {
+		return false
 	}
 
-	if opts.Internal && trcInfo.TraceType == madmin.TraceHTTP && HasPrefix(trcInfo.ReqInfo.Path, minioReservedBucketPath+SlashSeparator) {
-		return true
+	// Check internal path
+	isInternal := isHTTP && HasPrefix(trcInfo.HTTP.ReqInfo.Path, minioReservedBucketPath+SlashSeparator)
+	if isInternal && !opts.Internal {
+		return false
 	}
 
-	if opts.S3 && trcInfo.TraceType == madmin.TraceHTTP && !HasPrefix(trcInfo.ReqInfo.Path, minioReservedBucketPath+SlashSeparator) {
-		return true
+	// Filter non-errors.
+	if isHTTP && opts.OnlyErrors && trcInfo.HTTP.RespInfo.StatusCode < http.StatusBadRequest {
+		return false
 	}
 
-	if opts.Storage && trcInfo.TraceType == madmin.TraceStorage {
-		return true
-	}
-
-	return opts.OS && trcInfo.TraceType == madmin.TraceOS
+	return true
 }
 
 func extractTraceOptions(r *http.Request) (opts madmin.ServiceTraceOpts, err error) {
-	q := r.Form
-
-	opts.OnlyErrors = q.Get("err") == "true"
-	opts.S3 = q.Get("s3") == "true"
-	opts.Internal = q.Get("internal") == "true"
-	opts.Storage = q.Get("storage") == "true"
-	opts.OS = q.Get("os") == "true"
-
+	if err := opts.ParseParams(r); err != nil {
+		return opts, err
+	}
 	// Support deprecated 'all' query
-	if q.Get("all") == "true" {
+	if r.Form.Get("all") == "true" {
 		opts.S3 = true
 		opts.Internal = true
 		opts.Storage = true
 		opts.OS = true
-	}
-
-	if t := q.Get("threshold"); t != "" {
-		d, err := time.ParseDuration(t)
-		if err != nil {
-			return opts, err
-		}
-		opts.Threshold = d
+		// Older mc - cannot deal with more types...
 	}
 	return
 }
@@ -1400,20 +1515,21 @@ func (a adminAPIHandlers) TraceHandler(w http.ResponseWriter, r *http.Request) {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
 		return
 	}
-
 	setEventStreamHeaders(w)
 
 	// Trace Publisher and peer-trace-client uses nonblocking send and hence does not wait for slow receivers.
 	// Use buffered channel to take care of burst sends or slow w.Write()
-	traceCh := make(chan interface{}, 4000)
+	traceCh := make(chan pubsub.Maskable, 4000)
 
 	peers, _ := newPeerRestClients(globalEndpoints)
+	mask := pubsub.MaskFromMaskable(traceOpts.TraceTypes())
 
-	traceFn := func(entry interface{}) bool {
-		return mustTrace(entry, traceOpts)
-	}
-
-	err = globalTrace.Subscribe(traceCh, ctx.Done(), traceFn)
+	err = globalTrace.Subscribe(mask, traceCh, ctx.Done(), func(entry pubsub.Maskable) bool {
+		if e, ok := entry.(madmin.TraceInfo); ok {
+			return shouldTrace(e, traceOpts)
+		}
+		return false
+	})
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrSlowDown), r.URL)
 		return
@@ -1454,7 +1570,7 @@ func (a adminAPIHandlers) TraceHandler(w http.ResponseWriter, r *http.Request) {
 	}
 }
 
-// The handler sends console logs to the connected HTTP client.
+// The ConsoleLogHandler handler sends console logs to the connected HTTP client.
 func (a adminAPIHandlers) ConsoleLogHandler(w http.ResponseWriter, r *http.Request) {
 	ctx := newContext(r, w, "ConsoleLog")
 
@@ -1472,11 +1588,10 @@ func (a adminAPIHandlers) ConsoleLogHandler(w http.ResponseWriter, r *http.Reque
 		limitLines = 10
 	}
 
-	logKind := r.Form.Get("logType")
-	if logKind == "" {
-		logKind = string(logger.All)
+	logKind := madmin.LogKind(strings.ToUpper(r.Form.Get("logType"))).LogMask()
+	if logKind == 0 {
+		logKind = madmin.LogMaskAll
 	}
-	logKind = strings.ToUpper(logKind)
 
 	// Avoid reusing tcp connection if read timeout is hit
 	// This is needed to make r.Context().Done() work as
@@ -1485,7 +1600,7 @@ func (a adminAPIHandlers) ConsoleLogHandler(w http.ResponseWriter, r *http.Reque
 
 	setEventStreamHeaders(w)
 
-	logCh := make(chan interface{}, 4000)
+	logCh := make(chan pubsub.Maskable, 4000)
 
 	peers, _ := newPeerRestClients(globalEndpoints)
 
@@ -1551,7 +1666,7 @@ func (a adminAPIHandlers) KMSCreateKeyHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	if err := GlobalKMS.CreateKey(r.Form.Get("key-id")); err != nil {
+	if err := GlobalKMS.CreateKey(ctx, r.Form.Get("key-id")); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
@@ -1573,7 +1688,7 @@ func (a adminAPIHandlers) KMSStatusHandler(w http.ResponseWriter, r *http.Reques
 		return
 	}
 
-	stat, err := GlobalKMS.Stat()
+	stat, err := GlobalKMS.Stat(ctx)
 	if err != nil {
 		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInternalError), err.Error(), r.URL)
 		return
@@ -1612,7 +1727,7 @@ func (a adminAPIHandlers) KMSKeyStatusHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	stat, err := GlobalKMS.Stat()
+	stat, err := GlobalKMS.Stat(ctx)
 	if err != nil {
 		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInternalError), err.Error(), r.URL)
 		return
@@ -1628,7 +1743,7 @@ func (a adminAPIHandlers) KMSKeyStatusHandler(w http.ResponseWriter, r *http.Req
 
 	kmsContext := kms.Context{"MinIO admin API": "KMSKeyStatusHandler"} // Context for a test key operation
 	// 1. Generate a new key using the KMS.
-	key, err := GlobalKMS.GenerateKey(keyID, kmsContext)
+	key, err := GlobalKMS.GenerateKey(ctx, keyID, kmsContext)
 	if err != nil {
 		response.EncryptionErr = err.Error()
 		resp, err := json.Marshal(response)
@@ -1872,6 +1987,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	healthCtx, healthCancel := context.WithTimeout(lkctx.Context(), deadline)
 	defer healthCancel()
 
+	// Freeze all incoming S3 API calls before running speedtest.
+	globalNotificationSys.ServiceFreeze(ctx, true)
+
+	// unfreeze all incoming S3 API calls after speedtest.
+	defer globalNotificationSys.ServiceFreeze(ctx, false)
+
 	hostAnonymizer := createHostAnonymizer()
 	// anonAddr - Anonymizes hosts in given input string.
 	anonAddr := func(addr string) string {
@@ -2109,11 +2230,6 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteDrivePerfInfo := func() {
 		if query.Get(string(madmin.HealthDataTypePerfDrive)) == "true" {
-			// Freeze all incoming S3 API calls before running speedtest.
-			globalNotificationSys.ServiceFreeze(ctx, true)
-			// unfreeze all incoming S3 API calls after speedtest.
-			defer globalNotificationSys.ServiceFreeze(ctx, false)
-
 			opts := madmin.DriveSpeedTestOpts{
 				Serial:    false,
 				BlockSize: 4 * humanize.MiByte,
@@ -2134,11 +2250,13 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	getAndWriteObjPerfInfo := func() {
 		if query.Get(string(madmin.HealthDataTypePerfObj)) == "true" {
 			concurrent := 32
+
+			storageInfo, _ := objectAPI.StorageInfo(ctx)
+
 			size := 64 * humanize.MiByte
 			autotune := true
 
-			sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, objectAPI, concurrent, size, autotune)
-
+			sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, storageInfo, concurrent, size, autotune)
 			if !sufficientCapacity {
 				healthInfo.Perf.Error = capacityErrMsg
 				partialWrite(healthInfo)
@@ -2149,9 +2267,9 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 				autotune = false
 			}
 
-			bucketExists, err := makeObjectPerfBucket(ctx, objectAPI)
+			bucketExists, err := makeObjectPerfBucket(ctx, objectAPI, globalObjectPerfBucket)
 			if err != nil {
-				healthInfo.Perf.Error = "Could not make object perf bucket: " + err.Error()
+				healthInfo.Perf.Error = "Unable to create bucket: " + err.Error()
 				partialWrite(healthInfo)
 				return
 			}
@@ -2160,14 +2278,8 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 				defer deleteObjectPerfBucket(objectAPI)
 			}
 
-			// Freeze all incoming S3 API calls before running speedtest.
-			globalNotificationSys.ServiceFreeze(ctx, true)
-
-			// unfreeze all incoming S3 API calls after speedtest.
-			defer globalNotificationSys.ServiceFreeze(ctx, false)
-
 			opts := speedTestOpts{
-				throughputSize:   size,
+				objectSize:       size,
 				concurrencyStart: concurrent,
 				duration:         10 * time.Second,
 				autotune:         autotune,
@@ -2187,19 +2299,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 				return
 			}
 
-			nsLock := objectAPI.NewNSLock(minioMetaBucket, "netperf")
-			lkctx, err := nsLock.GetLock(ctx, globalOperationTimeout)
-			if err != nil {
-				healthInfo.Perf.Error = "Could not acquire lock for netperf: " + err.Error()
-			} else {
-				defer nsLock.Unlock(lkctx.Cancel)
-
-				netPerf := globalNotificationSys.Netperf(ctx, time.Second*10)
-				for _, np := range netPerf {
-					np.Endpoint = anonAddr(np.Endpoint)
-					healthInfo.Perf.NetPerf = append(healthInfo.Perf.NetPerf, np)
-				}
+			netPerf := globalNotificationSys.Netperf(ctx, time.Second*10)
+			for _, np := range netPerf {
+				np.Endpoint = anonAddr(np.Endpoint)
+				healthInfo.Perf.NetPerf = append(healthInfo.Perf.NetPerf, np)
 			}
+
 			partialWrite(healthInfo)
 		}
 	}
@@ -2486,7 +2591,7 @@ func fetchKMSStatus() madmin.KMS {
 		return kmsStat
 	}
 
-	stat, err := GlobalKMS.Stat()
+	stat, err := GlobalKMS.Stat(context.Background())
 	if err != nil {
 		kmsStat.Status = string(madmin.ItemOffline)
 		return kmsStat
@@ -2499,7 +2604,7 @@ func fetchKMSStatus() madmin.KMS {
 
 	kmsContext := kms.Context{"MinIO admin API": "ServerInfoHandler"} // Context for a test key operation
 	// 1. Generate a new key using the KMS.
-	key, err := GlobalKMS.GenerateKey("", kmsContext)
+	key, err := GlobalKMS.GenerateKey(context.Background(), "", kmsContext)
 	if err != nil {
 		kmsStat.Encrypt = fmt.Sprintf("Encryption failed: %v", err)
 	} else {
@@ -2580,6 +2685,100 @@ func checkConnection(endpointStr string, timeout time.Duration) error {
 	return nil
 }
 
+type clusterInfo struct {
+	DeploymentID string `json:"deployment_id"`
+	ClusterName  string `json:"cluster_name"`
+	UsedCapacity uint64 `json:"used_capacity"`
+	Info         struct {
+		MinioVersion    string `json:"minio_version"`
+		PoolsCount      int    `json:"pools_count"`
+		ServersCount    int    `json:"servers_count"`
+		DrivesCount     int    `json:"drives_count"`
+		BucketsCount    uint64 `json:"buckets_count"`
+		ObjectsCount    uint64 `json:"objects_count"`
+		TotalDriveSpace uint64 `json:"total_drive_space"`
+		UsedDriveSpace  uint64 `json:"used_drive_space"`
+	} `json:"info"`
+}
+
+func embedFileInZip(zipWriter *zip.Writer, name string, data []byte) error {
+	// Send profiling data to zip as file
+	header, zerr := zip.FileInfoHeader(dummyFileInfo{
+		name:    name,
+		size:    int64(len(data)),
+		mode:    0o600,
+		modTime: UTCNow(),
+		isDir:   false,
+		sys:     nil,
+	})
+	if zerr != nil {
+		return zerr
+	}
+	header.Method = zip.Deflate
+	zwriter, zerr := zipWriter.CreateHeader(header)
+	if zerr != nil {
+		return zerr
+	}
+	_, err := io.Copy(zwriter, bytes.NewReader(data))
+	return err
+}
+
+// appendClusterMetaInfoToZip gets information of the current cluster and embedded
+// it in the passed zipwriter, This is not a critical function and it is allowed
+// to fail with a ten seconds timeout.
+func appendClusterMetaInfoToZip(ctx context.Context, zipWriter *zip.Writer) {
+	objectAPI := newObjectLayerFn()
+	if objectAPI == nil {
+		return
+	}
+
+	// Add a ten seconds timeout because getting profiling data
+	// is critical for debugging, in contrary to getting cluster info
+	ctx, cancel := context.WithTimeout(ctx, 10*time.Second)
+	defer cancel()
+
+	resultCh := make(chan clusterInfo)
+	go func() {
+		ci := clusterInfo{}
+		ci.Info.PoolsCount = len(globalEndpoints)
+		ci.Info.ServersCount = len(globalEndpoints.Hostnames())
+		ci.Info.MinioVersion = Version
+
+		si, _ := objectAPI.StorageInfo(ctx)
+
+		ci.Info.DrivesCount = len(si.Disks)
+		for _, disk := range si.Disks {
+			ci.Info.TotalDriveSpace += disk.TotalSpace
+			ci.Info.UsedDriveSpace += disk.UsedSpace
+		}
+
+		dataUsageInfo, _ := loadDataUsageFromBackend(ctx, objectAPI)
+
+		ci.UsedCapacity = dataUsageInfo.ObjectsTotalSize
+		ci.Info.BucketsCount = dataUsageInfo.BucketsCount
+		ci.Info.ObjectsCount = dataUsageInfo.ObjectsTotalCount
+
+		ci.DeploymentID = globalDeploymentID
+		ci.ClusterName = fmt.Sprintf("%d-servers-%d-disks-%s", ci.Info.ServersCount, ci.Info.DrivesCount, ci.Info.MinioVersion)
+		resultCh <- ci
+	}()
+
+	select {
+	case <-ctx.Done():
+		return
+	case ci := <-resultCh:
+		out, err := json.MarshalIndent(ci, "", "   ")
+		if err != nil {
+			logger.LogIf(ctx, err)
+			return
+		}
+		err = embedFileInZip(zipWriter, "cluster.info", out)
+		if err != nil {
+			logger.LogIf(ctx, err)
+		}
+	}
+}
+
 // getRawDataer provides an interface for getting raw FS files.
 type getRawDataer interface {
 	GetRawData(ctx context.Context, volume, file string, fn func(r io.Reader, host string, disk string, filename string, info StatInfo) error) error
@@ -2599,7 +2798,8 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 	}
 	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
 
-	o, ok := newObjectLayerFn().(getRawDataer)
+	objLayer := newObjectLayerFn()
+	o, ok := objLayer.(getRawDataer)
 	if !ok {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
@@ -2722,6 +2922,8 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 	}); err != nil {
 		logger.LogIf(ctx, err)
 	}
+
+	appendClusterMetaInfoToZip(ctx, zipWriter)
 }
 
 func createHostAnonymizerForFSMode() map[string]string {

cmd/admin-heal-ops.go

@@ -91,8 +91,11 @@ type allHealState struct {
 	sync.RWMutex
 
 	// map of heal path to heal sequence
-	healSeqMap     map[string]*healSequence // Indexed by endpoint
-	healLocalDisks map[Endpoint]struct{}
+	healSeqMap map[string]*healSequence // Indexed by endpoint
+	// keep track of the healing status of disks in the memory
+	//   false: the disk needs to be healed but no healing routine is started
+	//    true: the disk is currently healing
+	healLocalDisks map[Endpoint]bool
 	healStatus     map[string]healingTracker // Indexed by disk ID
 }
 
@@ -100,7 +103,7 @@ type allHealState struct {
 func newHealState(cleanup bool) *allHealState {
 	hstate := &allHealState{
 		healSeqMap:     make(map[string]*healSequence),
-		healLocalDisks: map[Endpoint]struct{}{},
+		healLocalDisks: make(map[Endpoint]bool),
 		healStatus:     make(map[string]healingTracker),
 	}
 	if cleanup {
@@ -109,13 +112,6 @@ func newHealState(cleanup bool) *allHealState {
 	return hstate
 }
 
-func (ahs *allHealState) healDriveCount() int {
-	ahs.RLock()
-	defer ahs.RUnlock()
-
-	return len(ahs.healLocalDisks)
-}
-
 func (ahs *allHealState) popHealLocalDisks(healLocalDisks ...Endpoint) {
 	ahs.Lock()
 	defer ahs.Unlock()
@@ -165,23 +161,34 @@ func (ahs *allHealState) getLocalHealingDisks() map[string]madmin.HealingDisk {
 	return dst
 }
 
+// getHealLocalDiskEndpoints() returns the list of disks that need
+// to be healed but there is no healing routine in progress on them.
 func (ahs *allHealState) getHealLocalDiskEndpoints() Endpoints {
 	ahs.RLock()
 	defer ahs.RUnlock()
 
 	var endpoints Endpoints
-	for ep := range ahs.healLocalDisks {
-		endpoints = append(endpoints, ep)
+	for ep, healing := range ahs.healLocalDisks {
+		if !healing {
+			endpoints = append(endpoints, ep)
+		}
 	}
 	return endpoints
 }
 
+func (ahs *allHealState) markDiskForHealing(ep Endpoint) {
+	ahs.Lock()
+	defer ahs.Unlock()
+
+	ahs.healLocalDisks[ep] = true
+}
+
 func (ahs *allHealState) pushHealLocalDisks(healLocalDisks ...Endpoint) {
 	ahs.Lock()
 	defer ahs.Unlock()
 
 	for _, ep := range healLocalDisks {
-		ahs.healLocalDisks[ep] = struct{}{}
+		ahs.healLocalDisks[ep] = false
 	}
 }
 
@@ -804,16 +811,6 @@ func (h *healSequence) healMinioSysMeta(objAPI ObjectLayer, metaPrefix string) f
 	}
 }
 
-// healDiskFormat - heals format.json, return value indicates if a
-// failure error occurred.
-func (h *healSequence) healDiskFormat() error {
-	if h.isQuitting() {
-		return errHealStopSignalled
-	}
-
-	return h.queueHealTask(healSource{bucket: SlashSeparator}, madmin.HealItemMetadata)
-}
-
 // healBuckets - check for all buckets heal or just particular bucket.
 func (h *healSequence) healBuckets(objAPI ObjectLayer, bucketsOnly bool) error {
 	if h.isQuitting() {
@@ -825,7 +822,7 @@ func (h *healSequence) healBuckets(objAPI ObjectLayer, bucketsOnly bool) error {
 		return h.healBucket(objAPI, h.bucket, bucketsOnly)
 	}
 
-	buckets, err := objAPI.ListBuckets(h.ctx)
+	buckets, err := objAPI.ListBuckets(h.ctx, BucketOptions{})
 	if err != nil {
 		return errFnHealFromAPIErr(h.ctx, err)
 	}

cmd/admin-router.go

@@ -66,6 +66,8 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/storageinfo").HandlerFunc(gz(httpTraceAll(adminAPI.StorageInfoHandler)))
 		// DataUsageInfo operations
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/datausageinfo").HandlerFunc(gz(httpTraceAll(adminAPI.DataUsageInfoHandler)))
+		// Metrics operation
+		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/metrics").HandlerFunc(gz(httpTraceAll(adminAPI.MetricsHandler)))
 
 		if globalIsDistErasure || globalIsErasure {
 			// Heal operations
@@ -170,6 +172,19 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		// Set Group Status
 		adminRouter.Methods(http.MethodPut).Path(adminVersion+"/set-group-status").HandlerFunc(gz(httpTraceHdrs(adminAPI.SetGroupStatus))).Queries("group", "{group:.*}").Queries("status", "{status:.*}")
 
+		// Export IAM info to zipped file
+		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/export-iam").HandlerFunc(httpTraceHdrs(adminAPI.ExportIAM))
+
+		// Import IAM info
+		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/import-iam").HandlerFunc(httpTraceHdrs(adminAPI.ImportIAM))
+
+		// IDentity Provider configuration APIs
+		adminRouter.Methods(http.MethodPut).Path(adminVersion+"/idp-config").HandlerFunc(gz(httpTraceHdrs(adminAPI.SetIdentityProviderCfg))).Queries("type", "{type:.*}").Queries("name", "{name:.*}")
+		adminRouter.Methods(http.MethodGet).Path(adminVersion+"/idp-config").HandlerFunc(gz(httpTraceHdrs(adminAPI.GetIdentityProviderCfg))).Queries("type", "{type:.*}")
+		adminRouter.Methods(http.MethodDelete).Path(adminVersion+"/idp-config").HandlerFunc(gz(httpTraceHdrs(adminAPI.DeleteIdentityProviderCfg))).Queries("type", "{type:.*}").Queries("name", "{name:.*}")
+
+		// -- END IAM APIs --
+
 		// GetBucketQuotaConfig
 		adminRouter.Methods(http.MethodGet).Path(adminVersion+"/get-bucket-quota").HandlerFunc(
 			gz(httpTraceHdrs(adminAPI.GetBucketQuotaConfigHandler))).Queries("bucket", "{bucket:.*}")
@@ -187,6 +202,17 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		// RemoveRemoteTargetHandler
 		adminRouter.Methods(http.MethodDelete).Path(adminVersion+"/remove-remote-target").HandlerFunc(
 			gz(httpTraceHdrs(adminAPI.RemoveRemoteTargetHandler))).Queries("bucket", "{bucket:.*}", "arn", "{arn:.*}")
+		// ReplicationDiff - MinIO extension API
+		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/replication/diff").HandlerFunc(
+			gz(httpTraceHdrs(adminAPI.ReplicationDiffHandler))).Queries("bucket", "{bucket:.*}")
+
+		// Bucket migration operations
+		// ExportBucketMetaHandler
+		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/export-bucket-metadata").HandlerFunc(
+			gz(httpTraceHdrs(adminAPI.ExportBucketMetadataHandler)))
+		// ImportBucketMetaHandler
+		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/import-bucket-metadata").HandlerFunc(
+			gz(httpTraceHdrs(adminAPI.ImportBucketMetadataHandler)))
 
 		// Remote Tier management operations
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/tier").HandlerFunc(gz(httpTraceHdrs(adminAPI.AddTierHandler)))
@@ -221,8 +247,8 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 				Queries("paths", "{paths:.*}").HandlerFunc(gz(httpTraceHdrs(adminAPI.ForceUnlockHandler)))
 		}
 
-		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest").HandlerFunc(httpTraceHdrs(adminAPI.SpeedtestHandler))
-		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest/object").HandlerFunc(httpTraceHdrs(adminAPI.ObjectSpeedtestHandler))
+		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest").HandlerFunc(httpTraceHdrs(adminAPI.SpeedTestHandler))
+		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest/object").HandlerFunc(httpTraceHdrs(adminAPI.ObjectSpeedTestHandler))
 		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest/drive").HandlerFunc(httpTraceHdrs(adminAPI.DriveSpeedtestHandler))
 		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest/net").HandlerFunc(httpTraceHdrs(adminAPI.NetperfHandler))
 

cmd/api-errors.go

@@ -196,8 +196,9 @@ const (
 	ErrInvalidTagDirective
 	// Add new error codes here.
 
-	// SSE-S3 related API errors
+	// SSE-S3/SSE-KMS related API errors
 	ErrInvalidEncryptionMethod
+	ErrInvalidEncryptionKeyID
 
 	// Server-Side-Encryption (with Customer provided key) related API errors.
 	ErrInsecureSSECustomerRequest
@@ -267,6 +268,8 @@ const (
 	ErrAdminConfigNoQuorum
 	ErrAdminConfigTooLarge
 	ErrAdminConfigBadJSON
+	ErrAdminNoSuchConfigTarget
+	ErrAdminConfigEnvOverridden
 	ErrAdminConfigDuplicateKeys
 	ErrAdminCredentialsMismatch
 	ErrInsecureClientRequest
@@ -885,7 +888,7 @@ var errorCodes = errorCodeMap{
 	},
 	ErrReplicationRemoteConnectionError: {
 		Code:           "XMinioAdminReplicationRemoteConnectionError",
-		Description:    "Remote service connection error - please check remote service credentials and target bucket",
+		Description:    "Remote service connection error",
 		HTTPStatusCode: http.StatusNotFound,
 	},
 	ErrReplicationBandwidthLimitError: {
@@ -1070,6 +1073,11 @@ var errorCodes = errorCodeMap{
 		Description:    "The encryption method specified is not supported",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
+	ErrInvalidEncryptionKeyID: {
+		Code:           "InvalidRequest",
+		Description:    "The specified KMS KeyID contains unsupported characters",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrInsecureSSECustomerRequest: {
 		Code:           "InvalidRequest",
 		Description:    "Requests specifying Server Side Encryption with Customer provided keys must be made over a secure connection.",
@@ -1156,7 +1164,7 @@ var errorCodes = errorCodeMap{
 	// MinIO extensions.
 	ErrStorageFull: {
 		Code:           "XMinioStorageFull",
-		Description:    "Storage backend has reached its minimum free disk threshold. Please delete a few objects to proceed.",
+		Description:    "Storage backend has reached its minimum free drive threshold. Please delete a few objects to proceed.",
 		HTTPStatusCode: http.StatusInsufficientStorage,
 	},
 	ErrRequestBodyParse: {
@@ -1240,11 +1248,21 @@ var errorCodes = errorCodeMap{
 			maxEConfigJSONSize),
 		HTTPStatusCode: http.StatusBadRequest,
 	},
+	ErrAdminNoSuchConfigTarget: {
+		Code:           "XMinioAdminNoSuchConfigTarget",
+		Description:    "No such named configuration target exists",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrAdminConfigBadJSON: {
 		Code:           "XMinioAdminConfigBadJSON",
 		Description:    "JSON configuration provided is of incorrect format",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
+	ErrAdminConfigEnvOverridden: {
+		Code:           "XMinioAdminConfigEnvOverridden",
+		Description:    "Unable to update config via Admin API due to environment variable override",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrAdminConfigDuplicateKeys: {
 		Code:           "XMinioAdminConfigDuplicateKeys",
 		Description:    "JSON configuration provided has objects with duplicate keys",
@@ -1909,6 +1927,8 @@ func toAPIErrorCode(ctx context.Context, err error) (apiErr APIErrorCode) {
 		apiErr = ErrInvalidEncryptionParameters
 	case crypto.ErrInvalidEncryptionMethod:
 		apiErr = ErrInvalidEncryptionMethod
+	case crypto.ErrInvalidEncryptionKeyID:
+		apiErr = ErrInvalidEncryptionKeyID
 	case crypto.ErrInvalidCustomerAlgorithm:
 		apiErr = ErrInvalidSSECustomerAlgorithm
 	case crypto.ErrMissingCustomerKey:
@@ -2054,7 +2074,7 @@ func toAPIErrorCode(ctx context.Context, err error) (apiErr APIErrorCode) {
 		apiErr = ErrRemoteDestinationNotFoundError
 	case BucketRemoteTargetNotFound:
 		apiErr = ErrRemoteTargetNotFoundError
-	case BucketRemoteConnectionErr:
+	case RemoteTargetConnectionErr:
 		apiErr = ErrReplicationRemoteConnectionError
 	case BucketRemoteAlreadyExists:
 		apiErr = ErrBucketRemoteAlreadyExists
@@ -2231,7 +2251,7 @@ func toAPIError(ctx context.Context, err error) APIError {
 			}
 		case crypto.Error:
 			apiErr = APIError{
-				Code:           "XMinIOEncryptionError",
+				Code:           "XMinioEncryptionError",
 				Description:    e.Error(),
 				HTTPStatusCode: http.StatusBadRequest,
 			}

cmd/api-headers.go

@@ -199,5 +199,12 @@ func setObjectHeaders(w http.ResponseWriter, objInfo ObjectInfo, rs *HTTPRangeSp
 		lc.SetPredictionHeaders(w, objInfo.ToLifecycleOpts())
 	}
 
+	if v, ok := objInfo.UserDefined[ReservedMetadataPrefix+"compression"]; ok {
+		if i := strings.LastIndexByte(v, '/'); i >= 0 {
+			v = v[i+1:]
+		}
+		w.Header()[xhttp.MinIOCompressed] = []string{v}
+	}
+
 	return nil
 }

cmd/auth-handler.go

@@ -500,11 +500,18 @@ func isSupportedS3AuthType(aType authType) bool {
 func setAuthHandler(h http.Handler) http.Handler {
 	// handler for validating incoming authorization headers.
 	return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		tc, ok := r.Context().Value(contextTraceReqKey).(*traceCtxt)
+
 		aType := getRequestAuthType(r)
 		if aType == authTypeSigned || aType == authTypeSignedV2 || aType == authTypeStreamingSigned {
 			// Verify if date headers are set, if not reject the request
 			amzDate, errCode := parseAmzDateHeader(r)
 			if errCode != ErrNone {
+				if ok {
+					tc.funcName = "handler.Auth"
+					tc.responseRecorder.LogErrBody = true
+				}
+
 				// All our internal APIs are sensitive towards Date
 				// header, for all requests where Date header is not
 				// present we will reject such clients.
@@ -516,6 +523,11 @@ func setAuthHandler(h http.Handler) http.Handler {
 			// or in the future, reject request otherwise.
 			curTime := UTCNow()
 			if curTime.Sub(amzDate) > globalMaxSkewTime || amzDate.Sub(curTime) > globalMaxSkewTime {
+				if ok {
+					tc.funcName = "handler.Auth"
+					tc.responseRecorder.LogErrBody = true
+				}
+
 				writeErrorResponse(r.Context(), w, errorCodes.ToAPIErr(ErrRequestTimeTooSkewed), r.URL)
 				atomic.AddUint64(&globalHTTPStats.rejectedRequestsTime, 1)
 				return
@@ -525,6 +537,12 @@ func setAuthHandler(h http.Handler) http.Handler {
 			h.ServeHTTP(w, r)
 			return
 		}
+
+		if ok {
+			tc.funcName = "handler.Auth"
+			tc.responseRecorder.LogErrBody = true
+		}
+
 		writeErrorResponse(r.Context(), w, errorCodes.ToAPIErr(ErrSignatureVersionNotSupported), r.URL)
 		atomic.AddUint64(&globalHTTPStats.rejectedRequestsAuth, 1)
 	})

cmd/auth-handler_test.go

@@ -32,6 +32,12 @@ import (
 	iampolicy "github.com/minio/pkg/iam/policy"
 )
 
+type nullReader struct{}
+
+func (r *nullReader) Read(b []byte) (int, error) {
+	return len(b), nil
+}
+
 // Test get request auth type.
 func TestGetRequestAuthType(t *testing.T) {
 	type testCase struct {

cmd/background-heal-ops.go

@@ -22,6 +22,7 @@ import (
 	"runtime"
 
 	"github.com/minio/madmin-go"
+	"github.com/minio/minio/internal/pubsub"
 )
 
 // healTask represents what to heal along with options
@@ -49,10 +50,10 @@ type healRoutine struct {
 	workers int
 }
 
-func systemIO() int {
+func activeListeners() int {
 	// Bucket notification and http trace are not costly, it is okay to ignore them
 	// while counting the number of concurrent connections
-	return int(globalHTTPListen.NumSubscribers()) + int(globalTrace.NumSubscribers())
+	return int(globalHTTPListen.NumSubscribers(pubsub.MaskAll)) + int(globalTrace.NumSubscribers(pubsub.MaskAll))
 }
 
 func waitForLowHTTPReq() {
@@ -61,7 +62,7 @@ func waitForLowHTTPReq() {
 		currentIO = httpServer.GetRequestCount
 	}
 
-	globalHealConfig.Wait(currentIO, systemIO)
+	globalHealConfig.Wait(currentIO, activeListeners)
 }
 
 func initBackgroundHealing(ctx context.Context, objAPI ObjectLayer) {

cmd/background-newdisks-heal-ops.go

@@ -26,15 +26,12 @@ import (
 	"os"
 	"sort"
 	"strings"
-	"sync"
 	"time"
 
 	"github.com/dustin/go-humanize"
 	"github.com/minio/madmin-go"
 	"github.com/minio/minio-go/v7/pkg/set"
-	"github.com/minio/minio/internal/color"
 	"github.com/minio/minio/internal/logger"
-	"github.com/minio/pkg/console"
 )
 
 const (
@@ -90,14 +87,14 @@ type healingTracker struct {
 // The disk ID will be validated against the loaded one.
 func loadHealingTracker(ctx context.Context, disk StorageAPI) (*healingTracker, error) {
 	if disk == nil {
-		return nil, errors.New("loadHealingTracker: nil disk given")
+		return nil, errors.New("loadHealingTracker: nil drive given")
 	}
 	diskID, err := disk.GetDiskID()
 	if err != nil {
 		return nil, err
 	}
 	b, err := disk.ReadAll(ctx, minioMetaBucket,
-		pathJoin(bucketMetaPrefix, slashSeparator, healingTrackerFilename))
+		pathJoin(bucketMetaPrefix, healingTrackerFilename))
 	if err != nil {
 		return nil, err
 	}
@@ -107,7 +104,7 @@ func loadHealingTracker(ctx context.Context, disk StorageAPI) (*healingTracker,
 		return nil, err
 	}
 	if h.ID != diskID && h.ID != "" {
-		return nil, fmt.Errorf("loadHealingTracker: disk id mismatch expected %s, got %s", h.ID, diskID)
+		return nil, fmt.Errorf("loadHealingTracker: drive id mismatch expected %s, got %s", h.ID, diskID)
 	}
 	h.disk = disk
 	h.ID = diskID
@@ -132,7 +129,7 @@ func newHealingTracker(disk StorageAPI) *healingTracker {
 // If the tracker has been deleted an error is returned.
 func (h *healingTracker) update(ctx context.Context) error {
 	if h.disk.Healing() == nil {
-		return fmt.Errorf("healingTracker: disk %q is not marked as healing", h.ID)
+		return fmt.Errorf("healingTracker: drive %q is not marked as healing", h.ID)
 	}
 	if h.ID == "" || h.PoolIndex < 0 || h.SetIndex < 0 || h.DiskIndex < 0 {
 		h.ID, _ = h.disk.GetDiskID()
@@ -158,15 +155,19 @@ func (h *healingTracker) save(ctx context.Context) error {
 	}
 	globalBackgroundHealState.updateHealStatus(h)
 	return h.disk.WriteAll(ctx, minioMetaBucket,
-		pathJoin(bucketMetaPrefix, slashSeparator, healingTrackerFilename),
+		pathJoin(bucketMetaPrefix, healingTrackerFilename),
 		htrackerBytes)
 }
 
 // delete the tracker on disk.
 func (h *healingTracker) delete(ctx context.Context) error {
 	return h.disk.Delete(ctx, minioMetaBucket,
-		pathJoin(bucketMetaPrefix, slashSeparator, healingTrackerFilename),
-		false)
+		pathJoin(bucketMetaPrefix, healingTrackerFilename),
+		DeleteOptions{
+			Recursive: false,
+			Force:     false,
+		},
+	)
 }
 
 func (h *healingTracker) isHealed(bucket string) bool {
@@ -258,26 +259,9 @@ func initAutoHeal(ctx context.Context, objAPI ObjectLayer) {
 
 	initBackgroundHealing(ctx, objAPI) // start quick background healing
 
-	bgSeq := mustGetHealSequence(ctx)
-
 	globalBackgroundHealState.pushHealLocalDisks(getLocalDisksToHeal()...)
 
-	if drivesToHeal := globalBackgroundHealState.healDriveCount(); drivesToHeal > 0 {
-		logger.Info(fmt.Sprintf("Found drives to heal %d, waiting until %s to heal the content - use 'mc admin heal alias/ --verbose' to check the status",
-			drivesToHeal, defaultMonitorNewDiskInterval))
-
-		// Heal any disk format and metadata early, if possible.
-		// Start with format healing
-		if err := bgSeq.healDiskFormat(); err != nil {
-			if newObjectLayerFn() != nil {
-				// log only in situations, when object layer
-				// has fully initialized.
-				logger.LogIf(bgSeq.ctx, err)
-			}
-		}
-	}
-
-	go monitorLocalDisksAndHeal(ctx, z, bgSeq)
+	go monitorLocalDisksAndHeal(ctx, z)
 }
 
 func getLocalDisksToHeal() (disksToHeal Endpoints) {
@@ -299,10 +283,111 @@ func getLocalDisksToHeal() (disksToHeal Endpoints) {
 	return disksToHeal
 }
 
+var newDiskHealingTimeout = newDynamicTimeout(30*time.Second, 10*time.Second)
+
+func healFreshDisk(ctx context.Context, z *erasureServerPools, endpoint Endpoint) error {
+	logger.Info(fmt.Sprintf("Proceeding to heal '%s' - 'mc admin heal alias/ --verbose' to check the status.", endpoint))
+
+	disk, format, err := connectEndpoint(endpoint)
+	if err != nil {
+		return fmt.Errorf("Error: %w, %s", err, endpoint)
+	}
+
+	poolIdx := globalEndpoints.GetLocalPoolIdx(disk.Endpoint())
+	if poolIdx < 0 {
+		return fmt.Errorf("unexpected pool index (%d) found in %s", poolIdx, disk.Endpoint())
+	}
+
+	// Calculate the set index where the current endpoint belongs
+	z.serverPools[poolIdx].erasureDisksMu.RLock()
+	setIdx, _, err := findDiskIndex(z.serverPools[poolIdx].format, format)
+	z.serverPools[poolIdx].erasureDisksMu.RUnlock()
+	if err != nil {
+		return err
+	}
+	if setIdx < 0 {
+		return fmt.Errorf("unexpected set index (%d) found in %s", setIdx, disk.Endpoint())
+	}
+
+	// Prevent parallel erasure set healing
+	locker := z.NewNSLock(minioMetaBucket, fmt.Sprintf("new-drive-healing/%s/%d/%d", endpoint, poolIdx, setIdx))
+	lkctx, err := locker.GetLock(ctx, newDiskHealingTimeout)
+	if err != nil {
+		return err
+	}
+	ctx = lkctx.Context()
+	defer locker.Unlock(lkctx.Cancel)
+
+	buckets, _ := z.ListBuckets(ctx, BucketOptions{})
+	// Buckets data are dispersed in multiple zones/sets, make
+	// sure to heal all bucket metadata configuration.
+	buckets = append(buckets, BucketInfo{
+		Name: pathJoin(minioMetaBucket, minioConfigPrefix),
+	}, BucketInfo{
+		Name: pathJoin(minioMetaBucket, bucketMetaPrefix),
+	})
+
+	// Heal latest buckets first.
+	sort.Slice(buckets, func(i, j int) bool {
+		a, b := strings.HasPrefix(buckets[i].Name, minioMetaBucket), strings.HasPrefix(buckets[j].Name, minioMetaBucket)
+		if a != b {
+			return a
+		}
+		return buckets[i].Created.After(buckets[j].Created)
+	})
+
+	if serverDebugLog {
+		logger.Info("Healing drive '%v' on %s pool", disk, humanize.Ordinal(poolIdx+1))
+	}
+
+	// Load healing tracker in this disk
+	tracker, err := loadHealingTracker(ctx, disk)
+	if err != nil {
+		// So someone changed the drives underneath, healing tracker missing.
+		logger.LogIf(ctx, fmt.Errorf("Healing tracker missing on '%s', drive was swapped again on %s pool: %w",
+			disk, humanize.Ordinal(poolIdx+1), err))
+		tracker = newHealingTracker(disk)
+	}
+
+	// Load bucket totals
+	cache := dataUsageCache{}
+	if err := cache.load(ctx, z.serverPools[poolIdx].sets[setIdx], dataUsageCacheName); err == nil {
+		dataUsageInfo := cache.dui(dataUsageRoot, nil)
+		tracker.ObjectsTotalCount = dataUsageInfo.ObjectsTotalCount
+		tracker.ObjectsTotalSize = dataUsageInfo.ObjectsTotalSize
+	}
+
+	tracker.PoolIndex, tracker.SetIndex, tracker.DiskIndex = disk.GetDiskLoc()
+	tracker.setQueuedBuckets(buckets)
+	if err := tracker.save(ctx); err != nil {
+		return err
+	}
+
+	// Start or resume healing of this erasure set
+	if err = z.serverPools[poolIdx].sets[setIdx].healErasureSet(ctx, tracker.QueuedBuckets, tracker); err != nil {
+		return err
+	}
+
+	if tracker.ItemsFailed > 0 {
+		logger.Info("Healing drive '%s' failed (healed: %d, failed: %d).", disk, tracker.ItemsHealed, tracker.ItemsFailed)
+	} else {
+		logger.Info("Healing drive '%s' complete (healed: %d, failed: %d).", disk, tracker.ItemsHealed, tracker.ItemsFailed)
+	}
+
+	if serverDebugLog {
+		tracker.printTo(os.Stdout)
+		logger.Info("\n")
+	}
+
+	logger.LogIf(ctx, tracker.delete(ctx))
+
+	return nil
+}
+
 // monitorLocalDisksAndHeal - ensures that detected new disks are healed
 //  1. Only the concerned erasure set will be listed and healed
 //  2. Only the node hosting the disk is responsible to perform the heal
-func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq *healSequence) {
+func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools) {
 	// Perform automatic disk healing when a disk is replaced locally.
 	diskCheckTimer := time.NewTimer(defaultMonitorNewDiskInterval)
 	defer diskCheckTimer.Stop()
@@ -312,139 +397,34 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 		case <-ctx.Done():
 			return
 		case <-diskCheckTimer.C:
-			var erasureSetInPoolDisksToHeal []map[int][]StorageAPI
-
 			healDisks := globalBackgroundHealState.getHealLocalDiskEndpoints()
-			if len(healDisks) > 0 {
-				// Reformat disks
-				bgSeq.queueHealTask(healSource{bucket: SlashSeparator}, madmin.HealItemMetadata)
-
-				// Ensure that reformatting disks is finished
-				bgSeq.queueHealTask(healSource{bucket: nopHeal}, madmin.HealItemMetadata)
-
-				logger.Info(fmt.Sprintf("Found drives to heal %d, proceeding to heal - 'mc admin heal alias/ --verbose' to check the status.",
-					len(healDisks)))
-
-				erasureSetInPoolDisksToHeal = make([]map[int][]StorageAPI, len(z.serverPools))
-				for i := range z.serverPools {
-					erasureSetInPoolDisksToHeal[i] = map[int][]StorageAPI{}
-				}
+			if len(healDisks) == 0 {
+				// Reset for next interval.
+				diskCheckTimer.Reset(defaultMonitorNewDiskInterval)
+				continue
 			}
 
-			if serverDebugLog && len(healDisks) > 0 {
-				console.Debugf(color.Green("healDisk:")+" disk check timer fired, attempting to heal %d drives\n", len(healDisks))
+			// Reformat disks immediately
+			_, err := z.HealFormat(context.Background(), false)
+			if err != nil && !errors.Is(err, errNoHealRequired) {
+				logger.LogIf(ctx, err)
+				// Reset for next interval.
+				diskCheckTimer.Reset(defaultMonitorNewDiskInterval)
+				continue
 			}
 
-			// heal only if new disks found.
-			for _, endpoint := range healDisks {
-				disk, format, err := connectEndpoint(endpoint)
-				if err != nil {
-					printEndpointError(endpoint, err, true)
-					continue
-				}
-
-				poolIdx := globalEndpoints.GetLocalPoolIdx(disk.Endpoint())
-				if poolIdx < 0 {
-					continue
-				}
-
-				// Calculate the set index where the current endpoint belongs
-				z.serverPools[poolIdx].erasureDisksMu.RLock()
-				// Protect reading reference format.
-				setIndex, _, err := findDiskIndex(z.serverPools[poolIdx].format, format)
-				z.serverPools[poolIdx].erasureDisksMu.RUnlock()
-				if err != nil {
-					printEndpointError(endpoint, err, false)
-					continue
-				}
-
-				erasureSetInPoolDisksToHeal[poolIdx][setIndex] = append(erasureSetInPoolDisksToHeal[poolIdx][setIndex], disk)
-			}
-
-			buckets, _ := z.ListBuckets(ctx)
-
-			// Buckets data are dispersed in multiple zones/sets, make
-			// sure to heal all bucket metadata configuration.
-			buckets = append(buckets, BucketInfo{
-				Name: pathJoin(minioMetaBucket, minioConfigPrefix),
-			}, BucketInfo{
-				Name: pathJoin(minioMetaBucket, bucketMetaPrefix),
-			})
-
-			// Heal latest buckets first.
-			sort.Slice(buckets, func(i, j int) bool {
-				a, b := strings.HasPrefix(buckets[i].Name, minioMetaBucket), strings.HasPrefix(buckets[j].Name, minioMetaBucket)
-				if a != b {
-					return a
-				}
-				return buckets[i].Created.After(buckets[j].Created)
-			})
-
-			// TODO(klauspost): This will block until all heals are done,
-			// in the future this should be able to start healing other sets at once.
-			var wg sync.WaitGroup
-			for i, setMap := range erasureSetInPoolDisksToHeal {
-				i := i
-				for setIndex, disks := range setMap {
-					if len(disks) == 0 {
-						continue
+			for _, disk := range healDisks {
+				go func(disk Endpoint) {
+					globalBackgroundHealState.markDiskForHealing(disk)
+					err := healFreshDisk(ctx, z, disk)
+					if err != nil {
+						printEndpointError(disk, err, false)
+						return
 					}
-					wg.Add(1)
-					go func(setIndex int, disks []StorageAPI) {
-						defer wg.Done()
-						for _, disk := range disks {
-							if serverDebugLog {
-								logger.Info("Healing disk '%v' on %s pool", disk, humanize.Ordinal(i+1))
-							}
-
-							// So someone changed the drives underneath, healing tracker missing.
-							tracker, err := loadHealingTracker(ctx, disk)
-							if err != nil {
-								logger.LogIf(ctx, fmt.Errorf("Healing tracker missing on '%s', disk was swapped again on %s pool: %w",
-									disk, humanize.Ordinal(i+1), err))
-								tracker = newHealingTracker(disk)
-							}
-
-							// Load bucket totals
-							cache := dataUsageCache{}
-							if err := cache.load(ctx, z.serverPools[i].sets[setIndex], dataUsageCacheName); err == nil {
-								dataUsageInfo := cache.dui(dataUsageRoot, nil)
-								tracker.ObjectsTotalCount = dataUsageInfo.ObjectsTotalCount
-								tracker.ObjectsTotalSize = dataUsageInfo.ObjectsTotalSize
-							}
-
-							tracker.PoolIndex, tracker.SetIndex, tracker.DiskIndex = disk.GetDiskLoc()
-							tracker.setQueuedBuckets(buckets)
-							if err := tracker.save(ctx); err != nil {
-								logger.LogIf(ctx, err)
-								// Unable to write healing tracker, permission denied or some
-								// other unexpected error occurred. Proceed to look for new
-								// disks to be healed again, we cannot proceed further.
-								return
-							}
-
-							err = z.serverPools[i].sets[setIndex].healErasureSet(ctx, tracker.QueuedBuckets, tracker)
-							if err != nil {
-								logger.LogIf(ctx, err)
-								continue
-							}
-
-							if serverDebugLog {
-								logger.Info("Healing disk '%s' on %s pool, %s set complete", disk,
-									humanize.Ordinal(i+1), humanize.Ordinal(setIndex+1))
-								logger.Info("Summary:\n")
-								tracker.printTo(os.Stdout)
-								logger.Info("\n")
-							}
-							logger.LogIf(ctx, tracker.delete(ctx))
-
-							// Only upon success pop the healed disk.
-							globalBackgroundHealState.popHealLocalDisks(disk.Endpoint())
-						}
-					}(setIndex, disks)
-				}
+					// Only upon success pop the healed disk.
+					globalBackgroundHealState.popHealLocalDisks(disk)
+				}(disk)
 			}
-			wg.Wait()
 
 			// Reset for next interval.
 			diskCheckTimer.Reset(defaultMonitorNewDiskInterval)

cmd/benchmark-utils_test.go

@@ -35,7 +35,7 @@ func runPutObjectBenchmark(b *testing.B, obj ObjectLayer, objSize int) {
 	// obtains random bucket name.
 	bucket := getRandomBucketName()
 	// create bucket.
-	err = obj.MakeBucketWithLocation(context.Background(), bucket, BucketOptions{})
+	err = obj.MakeBucketWithLocation(context.Background(), bucket, MakeBucketOptions{})
 	if err != nil {
 		b.Fatal(err)
 	}
@@ -76,7 +76,7 @@ func runPutObjectPartBenchmark(b *testing.B, obj ObjectLayer, partSize int) {
 	object := getRandomObjectName()
 
 	// create bucket.
-	err = obj.MakeBucketWithLocation(context.Background(), bucket, BucketOptions{})
+	err = obj.MakeBucketWithLocation(context.Background(), bucket, MakeBucketOptions{})
 	if err != nil {
 		b.Fatal(err)
 	}
@@ -196,7 +196,7 @@ func runPutObjectBenchmarkParallel(b *testing.B, obj ObjectLayer, objSize int) {
 	// obtains random bucket name.
 	bucket := getRandomBucketName()
 	// create bucket.
-	err := obj.MakeBucketWithLocation(context.Background(), bucket, BucketOptions{})
+	err := obj.MakeBucketWithLocation(context.Background(), bucket, MakeBucketOptions{})
 	if err != nil {
 		b.Fatal(err)
 	}

cmd/bitrot-streaming.go

@@ -152,10 +152,10 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 		streamOffset := (offset/b.shardSize)*int64(b.h.Size()) + offset
 		if len(b.data) == 0 && b.tillOffset != streamOffset {
 			b.rc, err = b.disk.ReadFileStream(context.TODO(), b.volume, b.filePath, streamOffset, b.tillOffset-streamOffset)
-			if err != nil {
+			if err != nil && err != errDiskNotFound {
 				logger.LogIf(GlobalContext,
-					fmt.Errorf("Error(%w) reading erasure shards at (%s: %s/%s), will attempt to reconstruct if we have quorum",
-						err, b.disk, b.volume, b.filePath))
+					fmt.Errorf("Reading erasure shards at (%s: %s/%s) returned '%w', will attempt to reconstruct if we have quorum",
+						b.disk, b.volume, b.filePath, err))
 			}
 		} else {
 			b.rc = io.NewSectionReader(bytes.NewReader(b.data), streamOffset, b.tillOffset-streamOffset)
@@ -180,7 +180,7 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 	b.h.Write(buf)
 
 	if !bytes.Equal(b.h.Sum(nil), b.hashBytes) {
-		logger.LogIf(GlobalContext, fmt.Errorf("Disk: %s  -> %s/%s - content hash does not match - expected %s, got %s",
+		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s  -> %s/%s - content hash does not match - expected %s, got %s",
 			b.disk, b.volume, b.filePath, hex.EncodeToString(b.hashBytes), hex.EncodeToString(b.h.Sum(nil))))
 		return 0, errFileCorrupt
 	}

cmd/bitrot-whole.go

@@ -38,12 +38,12 @@ type wholeBitrotWriter struct {
 func (b *wholeBitrotWriter) Write(p []byte) (int, error) {
 	err := b.disk.AppendFile(context.TODO(), b.volume, b.filePath, p)
 	if err != nil {
-		logger.LogIf(GlobalContext, fmt.Errorf("Disk: %s returned %w", b.disk, err))
+		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s returned %w", b.disk, err))
 		return 0, err
 	}
 	_, err = b.Hash.Write(p)
 	if err != nil {
-		logger.LogIf(GlobalContext, fmt.Errorf("Disk: %s returned %w", b.disk, err))
+		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s returned %w", b.disk, err))
 		return 0, err
 	}
 	return len(p), nil
@@ -72,12 +72,12 @@ func (b *wholeBitrotReader) ReadAt(buf []byte, offset int64) (n int, err error)
 	if b.buf == nil {
 		b.buf = make([]byte, b.tillOffset-offset)
 		if _, err := b.disk.ReadFile(context.TODO(), b.volume, b.filePath, offset, b.buf, b.verifier); err != nil {
-			logger.LogIf(GlobalContext, fmt.Errorf("Disk: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, err))
+			logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, err))
 			return 0, err
 		}
 	}
 	if len(b.buf) < len(buf) {
-		logger.LogIf(GlobalContext, fmt.Errorf("Disk: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, errLessData))
+		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, errLessData))
 		return 0, errLessData
 	}
 	n = copy(buf, b.buf)

cmd/bitrot_test.go

@@ -20,17 +20,11 @@ package cmd
 import (
 	"context"
 	"io"
-	"io/ioutil"
-	"os"
 	"testing"
 )
 
 func testBitrotReaderWriterAlgo(t *testing.T, bitrotAlgo BitrotAlgorithm) {
-	tmpDir, err := ioutil.TempDir("", "")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(tmpDir)
+	tmpDir := t.TempDir()
 
 	volume := "testvol"
 	filePath := "testfile"
@@ -60,7 +54,9 @@ func testBitrotReaderWriterAlgo(t *testing.T, bitrotAlgo BitrotAlgorithm) {
 	if err != nil {
 		t.Fatal(err)
 	}
-	writer.(io.Closer).Close()
+	if bw, ok := writer.(io.Closer); ok {
+		bw.Close()
+	}
 
 	reader := newBitrotReader(disk, nil, volume, filePath, 35, bitrotAlgo, bitrotWriterSum(writer), 10)
 	b := make([]byte, 10)
@@ -76,6 +72,9 @@ func testBitrotReaderWriterAlgo(t *testing.T, bitrotAlgo BitrotAlgorithm) {
 	if _, err = reader.ReadAt(b[:5], 30); err != nil {
 		t.Fatal(err)
 	}
+	if br, ok := reader.(io.Closer); ok {
+		br.Close()
+	}
 }
 
 func TestAllBitrotAlgorithms(t *testing.T) {

cmd/bucket-encryption-handlers.go

@@ -65,7 +65,7 @@ func (api objectAPIHandlers) PutBucketEncryptionHandler(w http.ResponseWriter, r
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -90,7 +90,7 @@ func (api objectAPIHandlers) PutBucketEncryptionHandler(w http.ResponseWriter, r
 	kmsKey := encConfig.KeyID()
 	if kmsKey != "" {
 		kmsContext := kms.Context{"MinIO admin API": "ServerInfoHandler"} // Context for a test key operation
-		_, err := GlobalKMS.GenerateKey(kmsKey, kmsContext)
+		_, err := GlobalKMS.GenerateKey(ctx, kmsKey, kmsContext)
 		if err != nil {
 			if errors.Is(err, kes.ErrKeyNotFound) {
 				writeErrorResponse(ctx, w, toAPIError(ctx, errKMSKeyNotFound), r.URL)
@@ -108,7 +108,8 @@ func (api objectAPIHandlers) PutBucketEncryptionHandler(w http.ResponseWriter, r
 	}
 
 	// Store the bucket encryption configuration in the object layer
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketSSEConfig, configData); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketSSEConfig, configData)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -122,6 +123,7 @@ func (api objectAPIHandlers) PutBucketEncryptionHandler(w http.ResponseWriter, r
 		Type:      madmin.SRBucketMetaTypeSSEConfig,
 		Bucket:    bucket,
 		SSEConfig: &cfgStr,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -153,7 +155,7 @@ func (api objectAPIHandlers) GetBucketEncryptionHandler(w http.ResponseWriter, r
 
 	// Check if bucket exists
 	var err error
-	if _, err = objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err = objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -196,13 +198,14 @@ func (api objectAPIHandlers) DeleteBucketEncryptionHandler(w http.ResponseWriter
 
 	// Check if bucket exists
 	var err error
-	if _, err = objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err = objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
 	// Delete bucket encryption config from object layer
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketSSEConfig, nil); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketSSEConfig, nil)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -212,6 +215,7 @@ func (api objectAPIHandlers) DeleteBucketEncryptionHandler(w http.ResponseWriter
 		Type:      madmin.SRBucketMetaTypeSSEConfig,
 		Bucket:    bucket,
 		SSEConfig: nil,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return

cmd/bucket-handlers.go

@@ -201,7 +201,7 @@ func (api objectAPIHandlers) GetBucketLocationHandler(w http.ResponseWriter, r *
 
 	getBucketInfo := objectAPI.GetBucketInfo
 
-	if _, err := getBucketInfo(ctx, bucket); err != nil {
+	if _, err := getBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -332,7 +332,7 @@ func (api objectAPIHandlers) ListBucketsHandler(w http.ResponseWriter, r *http.R
 	} else {
 		// Invoke the list buckets.
 		var err error
-		bucketsInfo, err = listBuckets(ctx)
+		bucketsInfo, err = listBuckets(ctx, BucketOptions{})
 		if err != nil {
 			writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 			return
@@ -447,7 +447,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 	checkRequestAuthType(ctx, r, policy.DeleteObjectAction, bucket, "")
 
 	// Before proceeding validate if bucket exists.
-	_, err := objectAPI.GetBucketInfo(ctx, bucket)
+	_, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -525,7 +525,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 			VersionSuspended: vc.Suspended(),
 		}
 
-		if replicateDeletes || hasLockEnabled || !globalTierConfigMgr.Empty() {
+		if replicateDeletes || object.VersionID != "" && hasLockEnabled || !globalTierConfigMgr.Empty() {
 			if !globalTierConfigMgr.Empty() && object.VersionID == "" && opts.VersionSuspended {
 				opts.VersionID = nullVersionID
 			}
@@ -554,7 +554,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 				object.ReplicateDecisionStr = dsc.String()
 			}
 		}
-		if hasLockEnabled {
+		if object.VersionID != "" && hasLockEnabled {
 			if apiErrCode := enforceRetentionBypassForDelete(ctx, r, bucket, object, goi, gerr); apiErrCode != ErrNone {
 				apiErr := errorCodes.ToAPIErr(apiErrCode)
 				deleteResults[index].errInfo = DeleteError{
@@ -649,6 +649,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 			dv := DeletedObjectReplicationInfo{
 				DeletedObject: dobj,
 				Bucket:        bucket,
+				EventType:     ReplicateIncomingDelete,
 			}
 			scheduleReplicationDelete(ctx, dv, objectAPI)
 		}
@@ -726,11 +727,30 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 		}
 	}
 
-	if s3Error := checkRequestAuthType(ctx, r, policy.CreateBucketAction, bucket, ""); s3Error != ErrNone {
+	cred, owner, s3Error := checkRequestAuthTypeCredential(ctx, r, policy.CreateBucketAction, bucket, "")
+	if s3Error != ErrNone {
 		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
 		return
 	}
 
+	if objectLockEnabled {
+		// Creating a bucket with locking requires the user having more permissions
+		for _, action := range []iampolicy.Action{iampolicy.PutBucketObjectLockConfigurationAction, iampolicy.PutBucketVersioningAction} {
+			if !globalIAMSys.IsAllowed(iampolicy.Args{
+				AccountName:     cred.AccessKey,
+				Groups:          cred.Groups,
+				Action:          action,
+				ConditionValues: getConditionValues(r, "", cred.AccessKey, cred.Claims),
+				BucketName:      bucket,
+				IsOwner:         owner,
+				Claims:          cred.Claims,
+			}) {
+				writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
+				return
+			}
+		}
+	}
+
 	// Parse incoming location constraint.
 	location, s3Error := parseLocationConstraint(r)
 	if s3Error != ErrNone {
@@ -745,7 +765,7 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	opts := BucketOptions{
+	opts := MakeBucketOptions{
 		Location:    location,
 		LockEnabled: objectLockEnabled,
 		ForceCreate: forceCreate,
@@ -764,7 +784,11 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 				}
 
 				if err = globalDNSConfig.Put(bucket); err != nil {
-					objectAPI.DeleteBucket(context.Background(), bucket, DeleteBucketOptions{Force: false, NoRecreate: true})
+					objectAPI.DeleteBucket(context.Background(), bucket, DeleteBucketOptions{
+						Force:      false,
+						NoRecreate: true,
+						SRDeleteOp: getSRBucketDeleteOp(globalSiteReplicationSys.isEnabled()),
+					})
 					writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 					return
 				}
@@ -806,19 +830,14 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 	}
 
 	// Proceed to creating a bucket.
-	err := objectAPI.MakeBucketWithLocation(ctx, bucket, opts)
-	if _, ok := err.(BucketExists); ok {
-		// Though bucket exists locally, we send the site-replication
-		// hook to ensure all sites have this bucket. If the hook
-		// succeeds, the client will still receive a bucket exists
-		// message.
-		err2 := globalSiteReplicationSys.MakeBucketHook(ctx, bucket, opts)
-		if err2 != nil {
-			writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-			return
+	if err := objectAPI.MakeBucketWithLocation(ctx, bucket, opts); err != nil {
+		if _, ok := err.(BucketExists); ok {
+			// Though bucket exists locally, we send the site-replication
+			// hook to ensure all sites have this bucket. If the hook
+			// succeeds, the client will still receive a bucket exists
+			// message.
+			globalSiteReplicationSys.MakeBucketHook(ctx, bucket, opts)
 		}
-	}
-	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -827,8 +846,7 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 	globalNotificationSys.LoadBucketMetadata(GlobalContext, bucket)
 
 	// Call site replication hook
-	err = globalSiteReplicationSys.MakeBucketHook(ctx, bucket, opts)
-	if err != nil {
+	if err := globalSiteReplicationSys.MakeBucketHook(ctx, bucket, opts); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -870,7 +888,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		return
 	}
 
-	if _, ok := crypto.IsRequested(r.Header); !objectAPI.IsEncryptionSupported() && ok {
+	if crypto.Requested(r.Header) && !objectAPI.IsEncryptionSupported() {
 		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -1046,7 +1064,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		return
 	}
 	if objectAPI.IsEncryptionSupported() {
-		if _, ok := crypto.IsRequested(formValues); ok && !HasSuffix(object, SlashSeparator) { // handle SSE requests
+		if crypto.Requested(formValues) && !HasSuffix(object, SlashSeparator) { // handle SSE requests
 			if crypto.SSECopy.IsRequested(r.Header) {
 				writeErrorResponse(ctx, w, toAPIError(ctx, errInvalidEncryptionParameters), r.URL)
 				return
@@ -1072,7 +1090,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 					return
 				}
 			}
-			reader, objectEncryptionKey, err = newEncryptReader(hashReader, kind, keyID, key, bucket, object, metadata, kmsCtx)
+			reader, objectEncryptionKey, err = newEncryptReader(ctx, hashReader, kind, keyID, key, bucket, object, metadata, kmsCtx)
 			if err != nil {
 				writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 				return
@@ -1121,9 +1139,12 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		Host:         handlers.GetSourceIP(r),
 	})
 
-	if successRedirect != "" {
-		// Replace raw query params..
-		redirectURL.RawQuery = getRedirectPostRawQuery(objInfo)
+	if redirectURL != nil { // success_action_redirect is valid and set.
+		v := redirectURL.Query()
+		v.Add("bucket", objInfo.Bucket)
+		v.Add("key", objInfo.Name)
+		v.Add("etag", "\""+objInfo.ETag+"\"")
+		redirectURL.RawQuery = v.Encode()
 		writeRedirectSeeOther(w, redirectURL.String())
 		return
 	}
@@ -1167,7 +1188,7 @@ func (api objectAPIHandlers) GetBucketPolicyStatusHandler(w http.ResponseWriter,
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1230,7 +1251,7 @@ func (api objectAPIHandlers) HeadBucketHandler(w http.ResponseWriter, r *http.Re
 
 	getBucketInfo := objectAPI.GetBucketInfo
 
-	if _, err := getBucketInfo(ctx, bucket); err != nil {
+	if _, err := getBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponseHeadersOnly(w, toAPIError(ctx, err))
 		return
 	}
@@ -1307,7 +1328,10 @@ func (api objectAPIHandlers) DeleteBucketHandler(w http.ResponseWriter, r *http.
 	deleteBucket := objectAPI.DeleteBucket
 
 	// Attempt to delete bucket.
-	if err := deleteBucket(ctx, bucket, DeleteBucketOptions{Force: forceDelete}); err != nil {
+	if err := deleteBucket(ctx, bucket, DeleteBucketOptions{
+		Force:      forceDelete,
+		SRDeleteOp: getSRBucketDeleteOp(globalSiteReplicationSys.isEnabled()),
+	}); err != nil {
 		apiErr := toAPIError(ctx, err)
 		if _, ok := err.(BucketNotEmpty); ok {
 			if globalBucketVersioningSys.Enabled(bucket) || globalBucketVersioningSys.Suspended(bucket) {
@@ -1391,7 +1415,8 @@ func (api objectAPIHandlers) PutBucketObjectLockConfigHandler(w http.ResponseWri
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, objectLockConfig, configData); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, objectLockConfig, configData)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1405,6 +1430,7 @@ func (api objectAPIHandlers) PutBucketObjectLockConfigHandler(w http.ResponseWri
 		Type:             madmin.SRBucketMetaTypeObjectLockConfig,
 		Bucket:           bucket,
 		ObjectLockConfig: &cfgStr,
+		UpdatedAt:        updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -1472,7 +1498,7 @@ func (api objectAPIHandlers) PutBucketTaggingHandler(w http.ResponseWriter, r *h
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1496,7 +1522,8 @@ func (api objectAPIHandlers) PutBucketTaggingHandler(w http.ResponseWriter, r *h
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketTaggingConfig, configData); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketTaggingConfig, configData)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1507,9 +1534,10 @@ func (api objectAPIHandlers) PutBucketTaggingHandler(w http.ResponseWriter, r *h
 	// errors.
 	cfgStr := base64.StdEncoding.EncodeToString(configData)
 	if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
-		Type:   madmin.SRBucketMetaTypeTags,
-		Bucket: bucket,
-		Tags:   &cfgStr,
+		Type:      madmin.SRBucketMetaTypeTags,
+		Bucket:    bucket,
+		Tags:      &cfgStr,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -1578,14 +1606,16 @@ func (api objectAPIHandlers) DeleteBucketTaggingHandler(w http.ResponseWriter, r
 		return
 	}
 
-	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketTaggingConfig, nil); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketTaggingConfig, nil)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
 	if err := globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
-		Type:   madmin.SRBucketMetaTypeTags,
-		Bucket: bucket,
+		Type:      madmin.SRBucketMetaTypeTags,
+		Bucket:    bucket,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return

cmd/bucket-lifecycle-handlers.go

@@ -62,7 +62,7 @@ func (api objectAPIHandlers) PutBucketLifecycleHandler(w http.ResponseWriter, r
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -91,7 +91,7 @@ func (api objectAPIHandlers) PutBucketLifecycleHandler(w http.ResponseWriter, r
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketLifecycleConfig, configData); err != nil {
+	if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketLifecycleConfig, configData); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -121,7 +121,7 @@ func (api objectAPIHandlers) GetBucketLifecycleHandler(w http.ResponseWriter, r
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -163,12 +163,12 @@ func (api objectAPIHandlers) DeleteBucketLifecycleHandler(w http.ResponseWriter,
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
-	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketLifecycleConfig, nil); err != nil {
+	if _, err := globalBucketMetadataSys.Update(ctx, bucket, bucketLifecycleConfig, nil); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/bucket-lifecycle.go

@@ -560,7 +560,7 @@ func (r *RestoreObjectRequest) validate(ctx context.Context, objAPI ObjectLayer)
 	}
 	// Check if bucket exists.
 	if !r.OutputLocation.IsEmpty() {
-		if _, err := objAPI.GetBucketInfo(ctx, r.OutputLocation.S3.BucketName); err != nil {
+		if _, err := objAPI.GetBucketInfo(ctx, r.OutputLocation.S3.BucketName, BucketOptions{}); err != nil {
 			return err
 		}
 		if r.OutputLocation.S3.Prefix == "" {

cmd/bucket-listobjects-handlers.go

@@ -19,7 +19,6 @@ package cmd
 
 import (
 	"context"
-	"fmt"
 	"net/http"
 	"strconv"
 	"strings"
@@ -101,7 +100,6 @@ func (api objectAPIHandlers) ListObjectVersionsHandler(w http.ResponseWriter, r
 	}
 
 	if err = DecryptETags(ctx, GlobalKMS, listObjectVersionsInfo.Objects); err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Failed to decrypt ETag: %v", err)) // TODO(aead): Remove once we are confident that decryption does not fail accidentially
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -167,7 +165,6 @@ func (api objectAPIHandlers) ListObjectsV2MHandler(w http.ResponseWriter, r *htt
 	}
 
 	if err = DecryptETags(ctx, GlobalKMS, listObjectsV2Info.Objects); err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Failed to decrypt ETag: %v", err)) // TODO(aead): Remove once we are confident that decryption does not fail accidentially
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -246,7 +243,6 @@ func (api objectAPIHandlers) ListObjectsV2Handler(w http.ResponseWriter, r *http
 	}
 
 	if err = DecryptETags(ctx, GlobalKMS, listObjectsV2Info.Objects); err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Failed to decrypt ETag: %v", err)) // TODO(aead): Remove once we are confident that decryption does not fail accidentially
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -347,7 +343,6 @@ func (api objectAPIHandlers) ListObjectsV1Handler(w http.ResponseWriter, r *http
 	}
 
 	if err = DecryptETags(ctx, GlobalKMS, listObjectsInfo.Objects); err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Failed to decrypt ETag: %v", err)) // TODO(aead): Remove once we are confident that decryption does not fail accidentially
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/bucket-metadata-sys.go

@@ -75,28 +75,28 @@ func (sys *BucketMetadataSys) Set(bucket string, meta BucketMetadata) {
 
 // Update update bucket metadata for the specified config file.
 // The configData data should not be modified after being sent here.
-func (sys *BucketMetadataSys) Update(ctx context.Context, bucket string, configFile string, configData []byte) error {
+func (sys *BucketMetadataSys) Update(ctx context.Context, bucket string, configFile string, configData []byte) (updatedAt time.Time, err error) {
 	objAPI := newObjectLayerFn()
 	if objAPI == nil {
-		return errServerNotInitialized
+		return updatedAt, errServerNotInitialized
 	}
 
 	if globalIsGateway && globalGatewayName != NASBackendGateway {
 		if configFile == bucketPolicyConfig {
 			if configData == nil {
-				return objAPI.DeleteBucketPolicy(ctx, bucket)
+				return updatedAt, objAPI.DeleteBucketPolicy(ctx, bucket)
 			}
 			config, err := policy.ParseConfig(bytes.NewReader(configData), bucket)
 			if err != nil {
-				return err
+				return updatedAt, err
 			}
-			return objAPI.SetBucketPolicy(ctx, bucket, config)
+			return updatedAt, objAPI.SetBucketPolicy(ctx, bucket, config)
 		}
-		return NotImplemented{}
+		return updatedAt, NotImplemented{}
 	}
 
 	if bucket == minioMetaBucket {
-		return errInvalidArgument
+		return updatedAt, errInvalidArgument
 	}
 
 	meta, err := loadBucketMetadata(ctx, objAPI, bucket)
@@ -105,56 +105,56 @@ func (sys *BucketMetadataSys) Update(ctx context.Context, bucket string, configF
 			// Only single drive mode needs this fallback.
 			meta = newBucketMetadata(bucket)
 		} else {
-			return err
+			return updatedAt, err
 		}
 	}
-
+	updatedAt = UTCNow()
 	switch configFile {
 	case bucketPolicyConfig:
 		meta.PolicyConfigJSON = configData
-		meta.PolicyConfigUpdatedAt = UTCNow()
+		meta.PolicyConfigUpdatedAt = updatedAt
 	case bucketNotificationConfig:
 		meta.NotificationConfigXML = configData
 	case bucketLifecycleConfig:
 		meta.LifecycleConfigXML = configData
 	case bucketSSEConfig:
 		meta.EncryptionConfigXML = configData
-		meta.EncryptionConfigUpdatedAt = UTCNow()
+		meta.EncryptionConfigUpdatedAt = updatedAt
 	case bucketTaggingConfig:
 		meta.TaggingConfigXML = configData
-		meta.TaggingConfigUpdatedAt = UTCNow()
+		meta.TaggingConfigUpdatedAt = updatedAt
 	case bucketQuotaConfigFile:
 		meta.QuotaConfigJSON = configData
-		meta.QuotaConfigUpdatedAt = UTCNow()
+		meta.QuotaConfigUpdatedAt = updatedAt
 	case objectLockConfig:
 		meta.ObjectLockConfigXML = configData
-		meta.ObjectLockConfigUpdatedAt = UTCNow()
+		meta.ObjectLockConfigUpdatedAt = updatedAt
 	case bucketVersioningConfig:
 		meta.VersioningConfigXML = configData
-		meta.VersioningConfigUpdatedAt = UTCNow()
+		meta.VersioningConfigUpdatedAt = updatedAt
 	case bucketReplicationConfig:
 		meta.ReplicationConfigXML = configData
-		meta.ReplicationConfigUpdatedAt = UTCNow()
+		meta.ReplicationConfigUpdatedAt = updatedAt
 	case bucketTargetsFile:
-		meta.BucketTargetsConfigJSON, meta.BucketTargetsConfigMetaJSON, err = encryptBucketMetadata(meta.Name, configData, kms.Context{
+		meta.BucketTargetsConfigJSON, meta.BucketTargetsConfigMetaJSON, err = encryptBucketMetadata(ctx, meta.Name, configData, kms.Context{
 			bucket:            meta.Name,
 			bucketTargetsFile: bucketTargetsFile,
 		})
 		if err != nil {
-			return fmt.Errorf("Error encrypting bucket target metadata %w", err)
+			return updatedAt, fmt.Errorf("Error encrypting bucket target metadata %w", err)
 		}
 	default:
-		return fmt.Errorf("Unknown bucket %s metadata update requested %s", bucket, configFile)
+		return updatedAt, fmt.Errorf("Unknown bucket %s metadata update requested %s", bucket, configFile)
 	}
 
 	if err := meta.Save(ctx, objAPI); err != nil {
-		return err
+		return updatedAt, err
 	}
 
 	sys.Set(bucket, meta)
 	globalNotificationSys.LoadBucketMetadata(bgContext(ctx), bucket) // Do not use caller context here
 
-	return nil
+	return updatedAt, nil
 }
 
 // Get metadata for a bucket.
@@ -338,6 +338,9 @@ func (sys *BucketMetadataSys) GetPolicyConfig(bucket string) (*policy.Policy, ti
 func (sys *BucketMetadataSys) GetQuotaConfig(ctx context.Context, bucket string) (*madmin.BucketQuota, time.Time, error) {
 	meta, err := sys.GetConfig(ctx, bucket)
 	if err != nil {
+		if errors.Is(err, errConfigNotFound) {
+			return nil, time.Time{}, BucketQuotaConfigNotFound{Bucket: bucket}
+		}
 		return nil, time.Time{}, err
 	}
 	return meta.quotaConfig, meta.QuotaConfigUpdatedAt, nil

cmd/bucket-metadata.go

@@ -105,10 +105,8 @@ type BucketMetadata struct {
 
 // newBucketMetadata creates BucketMetadata with the supplied name and Created to Now.
 func newBucketMetadata(name string) BucketMetadata {
-	now := UTCNow()
 	return BucketMetadata{
-		Name:    name,
-		Created: now,
+		Name: name,
 		notificationConfig: &event.Config{
 			XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/",
 		},
@@ -121,6 +119,17 @@ func newBucketMetadata(name string) BucketMetadata {
 	}
 }
 
+// SetCreatedAt preserves the CreatedAt time for bucket across sites in site replication. It defaults to
+// creation time of bucket on this cluster in all other cases.
+func (b *BucketMetadata) SetCreatedAt(createdAt time.Time) {
+	if b.Created.IsZero() {
+		b.Created = UTCNow()
+	}
+	if !createdAt.IsZero() {
+		b.Created = createdAt.UTC()
+	}
+}
+
 // Load - loads the metadata of bucket by name from ObjectLayer api.
 // If an error is returned the returned metadata will be default initialized.
 func (b *BucketMetadata) Load(ctx context.Context, api ObjectLayer, name string) error {
@@ -160,17 +169,19 @@ func loadBucketMetadata(ctx context.Context, objectAPI ObjectLayer, bucket strin
 	if err != nil && !errors.Is(err, errConfigNotFound) {
 		return b, err
 	}
-
+	if err == nil {
+		b.defaultTimestamps()
+	}
 	// Old bucket without bucket metadata. Hence we migrate existing settings.
 	if err := b.convertLegacyConfigs(ctx, objectAPI); err != nil {
 		return b, err
 	}
 
 	// migrate unencrypted remote targets
-	if err = b.migrateTargetConfig(ctx, objectAPI); err != nil {
+	if err := b.migrateTargetConfig(ctx, objectAPI); err != nil {
 		return b, err
 	}
-	b.defaultTimestamps()
+
 	return b, nil
 }
 
@@ -345,6 +356,7 @@ func (b *BucketMetadata) convertLegacyConfigs(ctx context.Context, objectAPI Obj
 			b.BucketTargetsConfigJSON = configData
 		}
 	}
+	b.defaultTimestamps()
 
 	if err := b.Save(ctx, objectAPI); err != nil {
 		return err
@@ -438,7 +450,7 @@ func (b *BucketMetadata) migrateTargetConfig(ctx context.Context, objectAPI Obje
 		return nil
 	}
 
-	encBytes, metaBytes, err := encryptBucketMetadata(b.Name, b.BucketTargetsConfigJSON, kms.Context{b.Name: b.Name, bucketTargetsFile: bucketTargetsFile})
+	encBytes, metaBytes, err := encryptBucketMetadata(ctx, b.Name, b.BucketTargetsConfigJSON, kms.Context{b.Name: b.Name, bucketTargetsFile: bucketTargetsFile})
 	if err != nil {
 		return err
 	}
@@ -449,14 +461,14 @@ func (b *BucketMetadata) migrateTargetConfig(ctx context.Context, objectAPI Obje
 }
 
 // encrypt bucket metadata if kms is configured.
-func encryptBucketMetadata(bucket string, input []byte, kmsContext kms.Context) (output, metabytes []byte, err error) {
+func encryptBucketMetadata(ctx context.Context, bucket string, input []byte, kmsContext kms.Context) (output, metabytes []byte, err error) {
 	if GlobalKMS == nil {
 		output = input
 		return
 	}
 
 	metadata := make(map[string]string)
-	key, err := GlobalKMS.GenerateKey("", kmsContext)
+	key, err := GlobalKMS.GenerateKey(ctx, "", kmsContext)
 	if err != nil {
 		return
 	}
@@ -465,7 +477,7 @@ func encryptBucketMetadata(bucket string, input []byte, kmsContext kms.Context)
 	objectKey := crypto.GenerateKey(key.Plaintext, rand.Reader)
 	sealedKey := objectKey.Seal(key.Plaintext, crypto.GenerateIV(rand.Reader), crypto.S3.String(), bucket, "")
 	crypto.S3.CreateMetadata(metadata, key.KeyID, key.Ciphertext, sealedKey)
-	_, err = sio.Encrypt(outbuf, bytes.NewBuffer(input), sio.Config{Key: objectKey[:], MinVersion: sio.Version20, CipherSuites: fips.CipherSuitesDARE()})
+	_, err = sio.Encrypt(outbuf, bytes.NewBuffer(input), sio.Config{Key: objectKey[:], MinVersion: sio.Version20, CipherSuites: fips.DARECiphers()})
 	if err != nil {
 		return output, metabytes, err
 	}
@@ -495,6 +507,6 @@ func decryptBucketMetadata(input []byte, bucket string, meta map[string]string,
 	}
 
 	outbuf := bytes.NewBuffer(nil)
-	_, err = sio.Decrypt(outbuf, bytes.NewBuffer(input), sio.Config{Key: objectKey[:], MinVersion: sio.Version20, CipherSuites: fips.CipherSuitesDARE()})
+	_, err = sio.Decrypt(outbuf, bytes.NewBuffer(input), sio.Config{Key: objectKey[:], MinVersion: sio.Version20, CipherSuites: fips.DARECiphers()})
 	return outbuf.Bytes(), err
 }

cmd/bucket-notification-handlers.go

@@ -60,7 +60,7 @@ func (api objectAPIHandlers) GetBucketNotificationHandler(w http.ResponseWriter,
 		return
 	}
 
-	_, err := objAPI.GetBucketInfo(ctx, bucketName)
+	_, err := objAPI.GetBucketInfo(ctx, bucketName, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -132,7 +132,7 @@ func (api objectAPIHandlers) PutBucketNotificationHandler(w http.ResponseWriter,
 		return
 	}
 
-	_, err := objectAPI.GetBucketInfo(ctx, bucketName)
+	_, err := objectAPI.GetBucketInfo(ctx, bucketName, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -160,7 +160,7 @@ func (api objectAPIHandlers) PutBucketNotificationHandler(w http.ResponseWriter,
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucketName, bucketNotificationConfig, configData); err != nil {
+	if _, err = globalBucketMetadataSys.Update(ctx, bucketName, bucketNotificationConfig, configData); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/bucket-policy-handlers.go

@@ -61,7 +61,7 @@ func (api objectAPIHandlers) PutBucketPolicyHandler(w http.ResponseWriter, r *ht
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -103,16 +103,18 @@ func (api objectAPIHandlers) PutBucketPolicyHandler(w http.ResponseWriter, r *ht
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketPolicyConfig, configData); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketPolicyConfig, configData)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
 	// Call site replication hook.
 	if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
-		Type:   madmin.SRBucketMetaTypePolicy,
-		Bucket: bucket,
-		Policy: bucketPolicyBytes,
+		Type:      madmin.SRBucketMetaTypePolicy,
+		Bucket:    bucket,
+		Policy:    bucketPolicyBytes,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -143,20 +145,22 @@ func (api objectAPIHandlers) DeleteBucketPolicyHandler(w http.ResponseWriter, r
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
-	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketPolicyConfig, nil); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketPolicyConfig, nil)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
 
 	// Call site replication hook.
 	if err := globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
-		Type:   madmin.SRBucketMetaTypePolicy,
-		Bucket: bucket,
+		Type:      madmin.SRBucketMetaTypePolicy,
+		Bucket:    bucket,
+		UpdatedAt: updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -187,7 +191,7 @@ func (api objectAPIHandlers) GetBucketPolicyHandler(w http.ResponseWriter, r *ht
 	}
 
 	// Check if bucket exists.
-	if _, err := objAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/bucket-policy-handlers_test.go

@@ -128,7 +128,7 @@ func testCreateBucket(obj ObjectLayer, instanceType, bucketName string, apiRoute
 			defer wg.Done()
 			// Sync start.
 			<-start
-			if err := obj.MakeBucketWithLocation(GlobalContext, bucketName1, BucketOptions{}); err != nil {
+			if err := obj.MakeBucketWithLocation(GlobalContext, bucketName1, MakeBucketOptions{}); err != nil {
 				if _, ok := err.(BucketExists); !ok {
 					t.Logf("unexpected error: %T: %v", err, err)
 					return
@@ -163,7 +163,7 @@ func testPutBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string
 	credentials auth.Credentials, t *testing.T,
 ) {
 	bucketName1 := fmt.Sprintf("%s-1", bucketName)
-	if err := obj.MakeBucketWithLocation(GlobalContext, bucketName1, BucketOptions{}); err != nil {
+	if err := obj.MakeBucketWithLocation(GlobalContext, bucketName1, MakeBucketOptions{}); err != nil {
 		t.Fatal(err)
 	}
 

cmd/bucket-replication-handlers.go

@@ -55,7 +55,7 @@ func (api objectAPIHandlers) PutBucketReplicationConfigHandler(w http.ResponseWr
 		return
 	}
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -86,7 +86,7 @@ func (api objectAPIHandlers) PutBucketReplicationConfigHandler(w http.ResponseWr
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, configData); err != nil {
+	if _, err = globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, configData); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -118,7 +118,7 @@ func (api objectAPIHandlers) GetBucketReplicationConfigHandler(w http.ResponseWr
 		return
 	}
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -157,7 +157,7 @@ func (api objectAPIHandlers) DeleteBucketReplicationConfigHandler(w http.Respons
 		return
 	}
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -165,7 +165,7 @@ func (api objectAPIHandlers) DeleteBucketReplicationConfigHandler(w http.Respons
 		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationDenyEditError), r.URL)
 		return
 	}
-	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, nil); err != nil {
+	if _, err := globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, nil); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -198,10 +198,16 @@ func (api objectAPIHandlers) GetBucketReplicationMetricsHandler(w http.ResponseW
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
+
+	if _, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
 	var usageInfo BucketUsageInfo
 	dataUsageInfo, err := loadDataUsageFromBackend(ctx, objectAPI)
 	if err == nil && !dataUsageInfo.LastUpdate.IsZero() {
@@ -261,7 +267,7 @@ func (api objectAPIHandlers) ResetBucketReplicationStartHandler(w http.ResponseW
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -304,7 +310,7 @@ func (api objectAPIHandlers) ResetBucketReplicationStartHandler(w http.ResponseW
 	rinfo.Targets = append(rinfo.Targets, ResyncTarget{Arn: tgtArns[0], ResetID: target.ResetID})
 	if err = globalBucketTargetSys.SetTarget(ctx, bucket, &target, true); err != nil {
 		switch err.(type) {
-		case BucketRemoteConnectionErr:
+		case RemoteTargetConnectionErr:
 			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrReplicationRemoteConnectionError, err), r.URL)
 		default:
 			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
@@ -350,7 +356,7 @@ func (api objectAPIHandlers) ResetBucketReplicationStatusHandler(w http.Response
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/bucket-replication-utils.go

@@ -21,12 +21,14 @@ import (
 	"bytes"
 	"fmt"
 	"net/http"
+	"net/url"
 	"regexp"
 	"strconv"
 	"strings"
 	"sync"
 	"time"
 
+	"github.com/minio/madmin-go"
 	"github.com/minio/minio/internal/bucket/replication"
 	xhttp "github.com/minio/minio/internal/http"
 )
@@ -503,6 +505,8 @@ func getHealReplicateObjectInfo(objInfo ObjectInfo, rcfg replicationConfig) Repl
 	}
 	var dsc ReplicateDecision
 	var tgtStatuses map[string]replication.StatusType
+	var purgeStatuses map[string]VersionPurgeStatusType
+
 	if oi.DeleteMarker || !oi.VersionPurgeStatus.Empty() {
 		dsc = checkReplicateDelete(GlobalContext, oi.Bucket, ObjectToDelete{
 			ObjectV: ObjectV{
@@ -516,15 +520,17 @@ func getHealReplicateObjectInfo(objInfo ObjectInfo, rcfg replicationConfig) Repl
 		}, replication.HealReplicationType, ObjectOptions{}))
 	}
 	tgtStatuses = replicationStatusesMap(oi.ReplicationStatusInternal)
-
+	purgeStatuses = versionPurgeStatusesMap(oi.VersionPurgeStatusInternal)
 	existingObjResync := rcfg.Resync(GlobalContext, oi, &dsc, tgtStatuses)
-
+	tm, _ := time.Parse(time.RFC3339Nano, oi.UserDefined[ReservedMetadataPrefixLower+ReplicationTimestamp])
 	return ReplicateObjectInfo{
-		ObjectInfo:        oi,
-		OpType:            replication.HealReplicationType,
-		Dsc:               dsc,
-		ExistingObjResync: existingObjResync,
-		TargetStatuses:    tgtStatuses,
+		ObjectInfo:           oi,
+		OpType:               replication.HealReplicationType,
+		Dsc:                  dsc,
+		ExistingObjResync:    existingObjResync,
+		TargetStatuses:       tgtStatuses,
+		TargetPurgeStatuses:  purgeStatuses,
+		ReplicationTimestamp: tm,
 	}
 }
 
@@ -724,3 +730,39 @@ func parseSizeFromContentRange(h http.Header) (sz int64, err error) {
 	}
 	return int64(usz), nil
 }
+
+func extractReplicateDiffOpts(q url.Values) (opts madmin.ReplDiffOpts) {
+	opts.Verbose = q.Get("verbose") == "true"
+	opts.ARN = q.Get("arn")
+	opts.Prefix = q.Get("prefix")
+	return
+}
+
+const (
+	replicationMRFDir = bucketMetaPrefix + SlashSeparator + replicationDir + SlashSeparator + "mrf"
+	mrfMetaFormat     = 1
+	mrfMetaVersionV1  = 1
+	mrfMetaVersion    = mrfMetaVersionV1
+)
+
+// MRFReplicateEntry mrf entry to save to disk
+type MRFReplicateEntry struct {
+	Bucket    string `json:"bucket" msg:"b"`
+	Object    string `json:"object" msg:"o"`
+	versionID string `json:"-"`
+}
+
+// MRFReplicateEntries has the map of MRF entries to save to disk
+type MRFReplicateEntries struct {
+	Entries map[string]MRFReplicateEntry `json:"entries" msg:"e"`
+	Version int                          `json:"version" msg:"v"`
+}
+
+// ToMRFEntry returns the relevant info needed by MRF
+func (ri ReplicateObjectInfo) ToMRFEntry() MRFReplicateEntry {
+	return MRFReplicateEntry{
+		Bucket:    ri.Bucket,
+		Object:    ri.Name,
+		versionID: ri.VersionID,
+	}
+}

cmd/bucket-replication-utils_gen.go

@@ -260,6 +260,413 @@ func (z *BucketReplicationResyncStatus) Msgsize() (s int) {
 	return
 }
 
+// DecodeMsg implements msgp.Decodable
+func (z *MRFReplicateEntries) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "e":
+			var zb0002 uint32
+			zb0002, err = dc.ReadMapHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "Entries")
+				return
+			}
+			if z.Entries == nil {
+				z.Entries = make(map[string]MRFReplicateEntry, zb0002)
+			} else if len(z.Entries) > 0 {
+				for key := range z.Entries {
+					delete(z.Entries, key)
+				}
+			}
+			for zb0002 > 0 {
+				zb0002--
+				var za0001 string
+				var za0002 MRFReplicateEntry
+				za0001, err = dc.ReadString()
+				if err != nil {
+					err = msgp.WrapError(err, "Entries")
+					return
+				}
+				var zb0003 uint32
+				zb0003, err = dc.ReadMapHeader()
+				if err != nil {
+					err = msgp.WrapError(err, "Entries", za0001)
+					return
+				}
+				for zb0003 > 0 {
+					zb0003--
+					field, err = dc.ReadMapKeyPtr()
+					if err != nil {
+						err = msgp.WrapError(err, "Entries", za0001)
+						return
+					}
+					switch msgp.UnsafeString(field) {
+					case "b":
+						za0002.Bucket, err = dc.ReadString()
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001, "Bucket")
+							return
+						}
+					case "o":
+						za0002.Object, err = dc.ReadString()
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001, "Object")
+							return
+						}
+					default:
+						err = dc.Skip()
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001)
+							return
+						}
+					}
+				}
+				z.Entries[za0001] = za0002
+			}
+		case "v":
+			z.Version, err = dc.ReadInt()
+			if err != nil {
+				err = msgp.WrapError(err, "Version")
+				return
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *MRFReplicateEntries) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 2
+	// write "e"
+	err = en.Append(0x82, 0xa1, 0x65)
+	if err != nil {
+		return
+	}
+	err = en.WriteMapHeader(uint32(len(z.Entries)))
+	if err != nil {
+		err = msgp.WrapError(err, "Entries")
+		return
+	}
+	for za0001, za0002 := range z.Entries {
+		err = en.WriteString(za0001)
+		if err != nil {
+			err = msgp.WrapError(err, "Entries")
+			return
+		}
+		// map header, size 2
+		// write "b"
+		err = en.Append(0x82, 0xa1, 0x62)
+		if err != nil {
+			return
+		}
+		err = en.WriteString(za0002.Bucket)
+		if err != nil {
+			err = msgp.WrapError(err, "Entries", za0001, "Bucket")
+			return
+		}
+		// write "o"
+		err = en.Append(0xa1, 0x6f)
+		if err != nil {
+			return
+		}
+		err = en.WriteString(za0002.Object)
+		if err != nil {
+			err = msgp.WrapError(err, "Entries", za0001, "Object")
+			return
+		}
+	}
+	// write "v"
+	err = en.Append(0xa1, 0x76)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt(z.Version)
+	if err != nil {
+		err = msgp.WrapError(err, "Version")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *MRFReplicateEntries) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 2
+	// string "e"
+	o = append(o, 0x82, 0xa1, 0x65)
+	o = msgp.AppendMapHeader(o, uint32(len(z.Entries)))
+	for za0001, za0002 := range z.Entries {
+		o = msgp.AppendString(o, za0001)
+		// map header, size 2
+		// string "b"
+		o = append(o, 0x82, 0xa1, 0x62)
+		o = msgp.AppendString(o, za0002.Bucket)
+		// string "o"
+		o = append(o, 0xa1, 0x6f)
+		o = msgp.AppendString(o, za0002.Object)
+	}
+	// string "v"
+	o = append(o, 0xa1, 0x76)
+	o = msgp.AppendInt(o, z.Version)
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *MRFReplicateEntries) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "e":
+			var zb0002 uint32
+			zb0002, bts, err = msgp.ReadMapHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Entries")
+				return
+			}
+			if z.Entries == nil {
+				z.Entries = make(map[string]MRFReplicateEntry, zb0002)
+			} else if len(z.Entries) > 0 {
+				for key := range z.Entries {
+					delete(z.Entries, key)
+				}
+			}
+			for zb0002 > 0 {
+				var za0001 string
+				var za0002 MRFReplicateEntry
+				zb0002--
+				za0001, bts, err = msgp.ReadStringBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Entries")
+					return
+				}
+				var zb0003 uint32
+				zb0003, bts, err = msgp.ReadMapHeaderBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Entries", za0001)
+					return
+				}
+				for zb0003 > 0 {
+					zb0003--
+					field, bts, err = msgp.ReadMapKeyZC(bts)
+					if err != nil {
+						err = msgp.WrapError(err, "Entries", za0001)
+						return
+					}
+					switch msgp.UnsafeString(field) {
+					case "b":
+						za0002.Bucket, bts, err = msgp.ReadStringBytes(bts)
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001, "Bucket")
+							return
+						}
+					case "o":
+						za0002.Object, bts, err = msgp.ReadStringBytes(bts)
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001, "Object")
+							return
+						}
+					default:
+						bts, err = msgp.Skip(bts)
+						if err != nil {
+							err = msgp.WrapError(err, "Entries", za0001)
+							return
+						}
+					}
+				}
+				z.Entries[za0001] = za0002
+			}
+		case "v":
+			z.Version, bts, err = msgp.ReadIntBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Version")
+				return
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *MRFReplicateEntries) Msgsize() (s int) {
+	s = 1 + 2 + msgp.MapHeaderSize
+	if z.Entries != nil {
+		for za0001, za0002 := range z.Entries {
+			_ = za0002
+			s += msgp.StringPrefixSize + len(za0001) + 1 + 2 + msgp.StringPrefixSize + len(za0002.Bucket) + 2 + msgp.StringPrefixSize + len(za0002.Object)
+		}
+	}
+	s += 2 + msgp.IntSize
+	return
+}
+
+// DecodeMsg implements msgp.Decodable
+func (z *MRFReplicateEntry) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "b":
+			z.Bucket, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Bucket")
+				return
+			}
+		case "o":
+			z.Object, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Object")
+				return
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z MRFReplicateEntry) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 2
+	// write "b"
+	err = en.Append(0x82, 0xa1, 0x62)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Bucket)
+	if err != nil {
+		err = msgp.WrapError(err, "Bucket")
+		return
+	}
+	// write "o"
+	err = en.Append(0xa1, 0x6f)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Object)
+	if err != nil {
+		err = msgp.WrapError(err, "Object")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z MRFReplicateEntry) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 2
+	// string "b"
+	o = append(o, 0x82, 0xa1, 0x62)
+	o = msgp.AppendString(o, z.Bucket)
+	// string "o"
+	o = append(o, 0xa1, 0x6f)
+	o = msgp.AppendString(o, z.Object)
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *MRFReplicateEntry) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "b":
+			z.Bucket, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Bucket")
+				return
+			}
+		case "o":
+			z.Object, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Object")
+				return
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z MRFReplicateEntry) Msgsize() (s int) {
+	s = 1 + 2 + msgp.StringPrefixSize + len(z.Bucket) + 2 + msgp.StringPrefixSize + len(z.Object)
+	return
+}
+
 // DecodeMsg implements msgp.Decodable
 func (z *ReplicateDecision) DecodeMsg(dc *msgp.Reader) (err error) {
 	var field []byte

cmd/bucket-replication-utils_gen_test.go

@@ -122,6 +122,232 @@ func BenchmarkDecodeBucketReplicationResyncStatus(b *testing.B) {
 	}
 }
 
+func TestMarshalUnmarshalMRFReplicateEntries(t *testing.T) {
+	v := MRFReplicateEntries{}
+	bts, err := v.MarshalMsg(nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	left, err := v.UnmarshalMsg(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after UnmarshalMsg(): %q", len(left), left)
+	}
+
+	left, err = msgp.Skip(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after Skip(): %q", len(left), left)
+	}
+}
+
+func BenchmarkMarshalMsgMRFReplicateEntries(b *testing.B) {
+	v := MRFReplicateEntries{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgMRFReplicateEntries(b *testing.B) {
+	v := MRFReplicateEntries{}
+	bts := make([]byte, 0, v.Msgsize())
+	bts, _ = v.MarshalMsg(bts[0:0])
+	b.SetBytes(int64(len(bts)))
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		bts, _ = v.MarshalMsg(bts[0:0])
+	}
+}
+
+func BenchmarkUnmarshalMRFReplicateEntries(b *testing.B) {
+	v := MRFReplicateEntries{}
+	bts, _ := v.MarshalMsg(nil)
+	b.ReportAllocs()
+	b.SetBytes(int64(len(bts)))
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		_, err := v.UnmarshalMsg(bts)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
+func TestEncodeDecodeMRFReplicateEntries(t *testing.T) {
+	v := MRFReplicateEntries{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeMRFReplicateEntries Msgsize() is inaccurate")
+	}
+
+	vn := MRFReplicateEntries{}
+	err := msgp.Decode(&buf, &vn)
+	if err != nil {
+		t.Error(err)
+	}
+
+	buf.Reset()
+	msgp.Encode(&buf, &v)
+	err = msgp.NewReader(&buf).Skip()
+	if err != nil {
+		t.Error(err)
+	}
+}
+
+func BenchmarkEncodeMRFReplicateEntries(b *testing.B) {
+	v := MRFReplicateEntries{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+	b.SetBytes(int64(buf.Len()))
+	en := msgp.NewWriter(msgp.Nowhere)
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.EncodeMsg(en)
+	}
+	en.Flush()
+}
+
+func BenchmarkDecodeMRFReplicateEntries(b *testing.B) {
+	v := MRFReplicateEntries{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+	b.SetBytes(int64(buf.Len()))
+	rd := msgp.NewEndlessReader(buf.Bytes(), b)
+	dc := msgp.NewReader(rd)
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		err := v.DecodeMsg(dc)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
+func TestMarshalUnmarshalMRFReplicateEntry(t *testing.T) {
+	v := MRFReplicateEntry{}
+	bts, err := v.MarshalMsg(nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	left, err := v.UnmarshalMsg(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after UnmarshalMsg(): %q", len(left), left)
+	}
+
+	left, err = msgp.Skip(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after Skip(): %q", len(left), left)
+	}
+}
+
+func BenchmarkMarshalMsgMRFReplicateEntry(b *testing.B) {
+	v := MRFReplicateEntry{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgMRFReplicateEntry(b *testing.B) {
+	v := MRFReplicateEntry{}
+	bts := make([]byte, 0, v.Msgsize())
+	bts, _ = v.MarshalMsg(bts[0:0])
+	b.SetBytes(int64(len(bts)))
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		bts, _ = v.MarshalMsg(bts[0:0])
+	}
+}
+
+func BenchmarkUnmarshalMRFReplicateEntry(b *testing.B) {
+	v := MRFReplicateEntry{}
+	bts, _ := v.MarshalMsg(nil)
+	b.ReportAllocs()
+	b.SetBytes(int64(len(bts)))
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		_, err := v.UnmarshalMsg(bts)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
+func TestEncodeDecodeMRFReplicateEntry(t *testing.T) {
+	v := MRFReplicateEntry{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeMRFReplicateEntry Msgsize() is inaccurate")
+	}
+
+	vn := MRFReplicateEntry{}
+	err := msgp.Decode(&buf, &vn)
+	if err != nil {
+		t.Error(err)
+	}
+
+	buf.Reset()
+	msgp.Encode(&buf, &v)
+	err = msgp.NewReader(&buf).Skip()
+	if err != nil {
+		t.Error(err)
+	}
+}
+
+func BenchmarkEncodeMRFReplicateEntry(b *testing.B) {
+	v := MRFReplicateEntry{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+	b.SetBytes(int64(buf.Len()))
+	en := msgp.NewWriter(msgp.Nowhere)
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.EncodeMsg(en)
+	}
+	en.Flush()
+}
+
+func BenchmarkDecodeMRFReplicateEntry(b *testing.B) {
+	v := MRFReplicateEntry{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+	b.SetBytes(int64(buf.Len()))
+	rd := msgp.NewEndlessReader(buf.Bytes(), b)
+	dc := msgp.NewReader(rd)
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		err := v.DecodeMsg(dc)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
 func TestMarshalUnmarshalReplicateDecision(t *testing.T) {
 	v := ReplicateDecision{}
 	bts, err := v.MarshalMsg(nil)

cmd/bucket-replication.go

@@ -347,7 +347,7 @@ func checkReplicateDelete(ctx context.Context, bucket string, dobj ObjectToDelet
 // target cluster, the object version is marked deleted on the source and hidden from listing. It is permanently
 // deleted from the source when the VersionPurgeStatus changes to "Complete", i.e after replication succeeds
 // on target.
-func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, objectAPI ObjectLayer, trigger string) {
+func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, objectAPI ObjectLayer) {
 	var replicationStatus replication.StatusType
 	bucket := dobj.Bucket
 	versionID := dobj.DeleteMarkerVersionID
@@ -358,7 +358,7 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 	defer func() {
 		replStatus := string(replicationStatus)
 		auditLogInternal(context.Background(), bucket, dobj.ObjectName, AuditLogOptions{
-			Trigger:   trigger,
+			Event:     dobj.EventType,
 			APIName:   ReplicateDeleteAPI,
 			VersionID: versionID,
 			Status:    replStatus,
@@ -403,6 +403,7 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 	lk := objectAPI.NewNSLock(bucket, "/[replicate]/"+dobj.ObjectName)
 	lkctx, err := lk.GetLock(ctx, globalOperationTimeout)
 	if err != nil {
+		globalReplicationPool.queueMRFSave(dobj.ToMRFEntry())
 		logger.LogIf(ctx, fmt.Errorf("failed to get lock for object: %s bucket:%s arn:%s", dobj.ObjectName, bucket, rcfg.RoleArn))
 		sendEvent(eventArgs{
 			BucketName: bucket,
@@ -477,8 +478,13 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 	eventName := event.ObjectReplicationComplete
 	if replicationStatus == replication.Failed {
 		eventName = event.ObjectReplicationFailed
+		globalReplicationPool.queueMRFSave(dobj.ToMRFEntry())
 	}
 	drs := getReplicationState(rinfos, dobj.ReplicationState, dobj.VersionID)
+	if replicationStatus != prevStatus {
+		drs.ReplicationTimeStamp = UTCNow()
+	}
+
 	dobjInfo, err := objectAPI.DeleteObject(ctx, bucket, dobj.ObjectName, ObjectOptions{
 		VersionID:         versionID,
 		MTime:             dobj.DeleteMarkerMTime.Time,
@@ -530,7 +536,7 @@ func replicateDeleteToTarget(ctx context.Context, dobj DeletedObjectReplicationI
 	if dobj.VersionID != "" && rinfo.VersionPurgeStatus == Complete {
 		return
 	}
-	if tgt.IsOffline() {
+	if globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 		logger.LogIf(ctx, fmt.Errorf("remote target is offline for bucket:%s arn:%s", dobj.Bucket, tgt.ARN))
 		sendEvent(eventArgs{
 			BucketName: dobj.Bucket,
@@ -857,7 +863,7 @@ func getReplicationAction(oi1 ObjectInfo, oi2 minio.ObjectInfo, opType replicati
 
 // replicateObject replicates the specified version of the object to destination bucket
 // The source object is then updated to reflect the replication status.
-func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI ObjectLayer, trigger string) {
+func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI ObjectLayer) {
 	var replicationStatus replication.StatusType
 	defer func() {
 		if replicationStatus.Empty() {
@@ -868,7 +874,7 @@ func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI Obje
 			replicationStatus = ri.ReplicationStatus
 		}
 		auditLogInternal(ctx, ri.Bucket, ri.Name, AuditLogOptions{
-			Trigger:   trigger,
+			Event:     ri.EventType,
 			APIName:   ReplicateObjectAPI,
 			VersionID: ri.VersionID,
 			Status:    replicationStatus.String(),
@@ -906,6 +912,7 @@ func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI Obje
 			Object:     objInfo,
 			Host:       "Internal: [Replication]",
 		})
+		globalReplicationPool.queueMRFSave(ri.ToMRFEntry())
 		logger.LogIf(ctx, fmt.Errorf("failed to get lock for object: %s bucket:%s arn:%s", object, bucket, cfg.RoleArn))
 		return
 	}
@@ -988,11 +995,12 @@ func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI Obje
 
 	// re-queue failures once more - keep a retry count to avoid flooding the queue if
 	// the target site is down. Leave it to scanner to catch up instead.
-	if rinfos.ReplicationStatus() != replication.Completed && ri.RetryCount < 1 {
+	if rinfos.ReplicationStatus() != replication.Completed {
 		ri.OpType = replication.HealReplicationType
+		ri.EventType = ReplicateMRF
 		ri.ReplicationStatusInternal = rinfos.ReplicationStatusInternal()
 		ri.RetryCount++
-		globalReplicationPool.queueReplicaFailedTask(ri)
+		globalReplicationPool.queueMRFSave(ri.ToMRFEntry())
 	}
 }
 
@@ -1034,7 +1042,7 @@ func replicateObjectToTarget(ctx context.Context, ri ReplicateObjectInfo, object
 		rinfo.ReplicationResynced = true
 		return
 	}
-	if tgt.IsOffline() {
+	if globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 		logger.LogIf(ctx, fmt.Errorf("remote target is offline for bucket:%s arn:%s", bucket, tgt.ARN))
 		sendEvent(eventArgs{
 			EventName:  event.ObjectReplicationNotTracked,
@@ -1054,13 +1062,15 @@ func replicateObjectToTarget(ctx context.Context, ri ReplicateObjectInfo, object
 		VersionSuspended: versionSuspended,
 	})
 	if err != nil {
-		sendEvent(eventArgs{
-			EventName:  event.ObjectReplicationNotTracked,
-			BucketName: bucket,
-			Object:     objInfo,
-			Host:       "Internal: [Replication]",
-		})
-		logger.LogIf(ctx, fmt.Errorf("Unable to update replicate for %s/%s(%s): %w", bucket, object, objInfo.VersionID, err))
+		if !isErrObjectNotFound(err) {
+			sendEvent(eventArgs{
+				EventName:  event.ObjectReplicationNotTracked,
+				BucketName: bucket,
+				Object:     objInfo,
+				Host:       "Internal: [Replication]",
+			})
+			logger.LogIf(ctx, fmt.Errorf("Unable to update replicate metadata for %s/%s(%s): %w", bucket, object, objInfo.VersionID, err))
+		}
 		return
 	}
 	defer func() {
@@ -1291,11 +1301,25 @@ func filterReplicationStatusMetadata(metadata map[string]string) map[string]stri
 type DeletedObjectReplicationInfo struct {
 	DeletedObject
 	Bucket    string
+	EventType string
 	OpType    replication.Type
 	ResetID   string
 	TargetArn string
 }
 
+// ToMRFEntry returns the relevant info needed by MRF
+func (di DeletedObjectReplicationInfo) ToMRFEntry() MRFReplicateEntry {
+	versionID := di.DeleteMarkerVersionID
+	if versionID == "" {
+		versionID = di.VersionID
+	}
+	return MRFReplicateEntry{
+		Bucket:    di.Bucket,
+		Object:    di.ObjectName,
+		versionID: versionID,
+	}
+}
+
 // Replication specific APIName
 const (
 	ReplicateObjectAPI = "ReplicateObject"
@@ -1313,12 +1337,15 @@ const (
 
 	// ReplicateMRF - audit trail for replication from Most Recent Failures (MRF) queue
 	ReplicateMRF = "replicate:mrf"
-	// ReplicateIncoming - audit trail indicating replication started [could be from incoming/existing/heal activity]
+	// ReplicateIncoming - audit trail of inline replication
 	ReplicateIncoming = "replicate:incoming"
+	// ReplicateIncomingDelete - audit trail of inline replication of deletes.
+	ReplicateIncomingDelete = "replicate:incoming:delete"
+
 	// ReplicateHeal - audit trail for healing of failed/pending replications
 	ReplicateHeal = "replicate:heal"
-	// ReplicateDelete - audit trail for delete replication
-	ReplicateDelete = "replicate:delete"
+	// ReplicateHealDelete - audit trail of healing of failed/pending delete replications.
+	ReplicateHealDelete = "replicate:heal:delete"
 )
 
 var (
@@ -1337,13 +1364,16 @@ type ReplicationPool struct {
 	mrfReplicaCh            chan ReplicateObjectInfo
 	existingReplicaCh       chan ReplicateObjectInfo
 	existingReplicaDeleteCh chan DeletedObjectReplicationInfo
-	workerSize              int
-	mrfWorkerSize           int
-	resyncState             replicationResyncState
-	workerWg                sync.WaitGroup
-	mrfWorkerWg             sync.WaitGroup
-	once                    sync.Once
-	mu                      sync.Mutex
+	mrfSaveCh               chan MRFReplicateEntry
+
+	workerSize    int
+	mrfWorkerSize int
+	resyncState   replicationResyncState
+	workerWg      sync.WaitGroup
+	mrfWorkerWg   sync.WaitGroup
+	once          sync.Once
+	mu            sync.Mutex
+	mrfMutex      sync.Mutex
 }
 
 // NewReplicationPool creates a pool of replication workers of specified size
@@ -1357,6 +1387,7 @@ func NewReplicationPool(ctx context.Context, o ObjectLayer, opts replicationPool
 		existingReplicaCh:       make(chan ReplicateObjectInfo, 100000),
 		existingReplicaDeleteCh: make(chan DeletedObjectReplicationInfo, 100000),
 		resyncState:             replicationResyncState{statusMap: make(map[string]BucketReplicationResyncStatus)},
+		mrfSaveCh:               make(chan MRFReplicateEntry, 100000),
 		ctx:                     ctx,
 		objLayer:                o,
 	}
@@ -1364,13 +1395,16 @@ func NewReplicationPool(ctx context.Context, o ObjectLayer, opts replicationPool
 	pool.ResizeWorkers(opts.Workers)
 	pool.ResizeFailedWorkers(opts.FailedWorkers)
 	go pool.AddExistingObjectReplicateWorker()
-	go pool.periodicResyncMetaSave(ctx, o)
+	go pool.updateResyncStatus(ctx, o)
+	go pool.processMRF()
+	go pool.persistMRF()
 	return pool
 }
 
 // AddMRFWorker adds a pending/failed replication worker to handle requests that could not be queued
 // to the other workers
 func (p *ReplicationPool) AddMRFWorker() {
+	defer p.mrfWorkerWg.Done()
 	for {
 		select {
 		case <-p.ctx.Done():
@@ -1379,7 +1413,7 @@ func (p *ReplicationPool) AddMRFWorker() {
 			if !ok {
 				return
 			}
-			replicateObject(p.ctx, oi, p.objLayer, ReplicateMRF)
+			replicateObject(p.ctx, oi, p.objLayer)
 		case <-p.mrfWorkerKillCh:
 			return
 		}
@@ -1397,12 +1431,12 @@ func (p *ReplicationPool) AddWorker() {
 			if !ok {
 				return
 			}
-			replicateObject(p.ctx, oi, p.objLayer, ReplicateIncoming)
+			replicateObject(p.ctx, oi, p.objLayer)
 		case doi, ok := <-p.replicaDeleteCh:
 			if !ok {
 				return
 			}
-			replicateDelete(p.ctx, doi, p.objLayer, ReplicateDelete)
+			replicateDelete(p.ctx, doi, p.objLayer)
 		case <-p.workerKillCh:
 			return
 		}
@@ -1419,12 +1453,12 @@ func (p *ReplicationPool) AddExistingObjectReplicateWorker() {
 			if !ok {
 				return
 			}
-			replicateObject(p.ctx, oi, p.objLayer, ReplicateExisting)
+			replicateObject(p.ctx, oi, p.objLayer)
 		case doi, ok := <-p.existingReplicaDeleteCh:
 			if !ok {
 				return
 			}
-			replicateDelete(p.ctx, doi, p.objLayer, ReplicateExistingDelete)
+			replicateDelete(p.ctx, doi, p.objLayer)
 		}
 	}
 }
@@ -1469,33 +1503,17 @@ func (p *ReplicationPool) suggestedWorkers(failQueue bool) int {
 	return int(float64(p.workerSize) * ReplicationWorkerMultiplier)
 }
 
-func (p *ReplicationPool) queueReplicaFailedTask(ri ReplicateObjectInfo) {
-	if p == nil {
-		return
-	}
-	select {
-	case <-GlobalContext.Done():
-		p.once.Do(func() {
-			close(p.replicaCh)
-			close(p.mrfReplicaCh)
-			close(p.existingReplicaCh)
-		})
-	case p.mrfReplicaCh <- ri:
-	default:
-		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Replication failed workers could not keep up with healing failures - consider increasing number of replication failed workers with `mc admin config set api replication_failed_workers=%d`", p.suggestedWorkers(true)), replicationSubsystem)
-	}
-}
-
 func (p *ReplicationPool) queueReplicaTask(ri ReplicateObjectInfo) {
 	if p == nil {
 		return
 	}
-	var ch chan ReplicateObjectInfo
+	var ch, healCh chan ReplicateObjectInfo
 	switch ri.OpType {
 	case replication.ExistingObjectReplicationType:
 		ch = p.existingReplicaCh
 	case replication.HealReplicationType:
-		fallthrough
+		ch = p.mrfReplicaCh
+		healCh = p.replicaCh
 	default:
 		ch = p.replicaCh
 	}
@@ -1506,9 +1524,10 @@ func (p *ReplicationPool) queueReplicaTask(ri ReplicateObjectInfo) {
 			close(p.mrfReplicaCh)
 			close(p.existingReplicaCh)
 		})
+	case healCh <- ri:
 	case ch <- ri:
 	default:
-		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Replication workers could not keep up with incoming traffic - consider increasing number of replication workers with `mc admin config set api replication_workers=%d`", p.suggestedWorkers(false)), replicationSubsystem)
+		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Unable to keep up with incoming traffic - we recommend increasing number of replicate object workers with `mc admin config set api replication_workers=%d`", p.suggestedWorkers(false)), string(replicationSubsystem))
 	}
 }
 
@@ -1545,7 +1564,7 @@ func (p *ReplicationPool) queueReplicaDeleteTask(doi DeletedObjectReplicationInf
 		})
 	case ch <- doi:
 	default:
-		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Replication workers could not keep up with incoming traffic - consider increasing number of replication workers with `mc admin config set api replication_workers=%d`", p.suggestedWorkers(false)), replicationSubsystem)
+		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Unable to keep up with incoming deletes - we recommend increasing number of replicate workers with `mc admin config set api replication_workers=%d`", p.suggestedWorkers(false)), string(replicationSubsystem))
 	}
 }
 
@@ -1649,7 +1668,7 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRa
 	}
 	for _, t := range proxyTargets.Targets {
 		tgt = globalBucketTargetSys.GetRemoteTargetClient(ctx, t.Arn)
-		if tgt == nil || tgt.IsOffline() {
+		if tgt == nil || globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 			continue
 		}
 		// if proxying explicitly disabled on remote target
@@ -1722,10 +1741,11 @@ func proxyHeadToReplicationTarget(ctx context.Context, bucket, object string, rs
 }
 
 func scheduleReplication(ctx context.Context, objInfo ObjectInfo, o ObjectLayer, dsc ReplicateDecision, opType replication.Type) {
+	ri := ReplicateObjectInfo{ObjectInfo: objInfo, OpType: opType, Dsc: dsc, EventType: ReplicateIncoming}
 	if dsc.Synchronous() {
-		replicateObject(ctx, ReplicateObjectInfo{ObjectInfo: objInfo, OpType: opType, Dsc: dsc}, o, ReplicateIncoming)
+		replicateObject(ctx, ri, o)
 	} else {
-		globalReplicationPool.queueReplicaTask(ReplicateObjectInfo{ObjectInfo: objInfo, OpType: opType, Dsc: dsc})
+		globalReplicationPool.queueReplicaTask(ri)
 	}
 	if sz, err := objInfo.GetActualSize(); err == nil {
 		for arn := range dsc.targetsMap {
@@ -1794,6 +1814,7 @@ func (c replicationConfig) Resync(ctx context.Context, oi ObjectInfo, dsc *Repli
 	objInfo.VersionPurgeStatusInternal = ""
 	objInfo.ReplicationStatus = ""
 	objInfo.VersionPurgeStatus = ""
+	delete(objInfo.UserDefined, xhttp.AmzBucketReplicationStatus)
 	resyncdsc := mustReplicate(ctx, oi.Bucket, oi.Name, getMustReplicateOptions(objInfo, replication.ExistingObjectReplicationType, ObjectOptions{}))
 	dsc = &resyncdsc
 	return c.resync(oi, dsc, tgtStatuses)
@@ -1956,10 +1977,10 @@ func getLatestReplicationStats(bucket string, u BucketUsageInfo) (s BucketReplic
 	return calculateBucketReplicationStats(bucket, u, bucketStats)
 }
 
-const resyncTimeInterval = time.Minute * 10
+const resyncTimeInterval = time.Minute * 1
 
-// periodicResyncMetaSave saves in-memory resync meta stats to disk in periodic intervals
-func (p *ReplicationPool) periodicResyncMetaSave(ctx context.Context, objectAPI ObjectLayer) {
+// updateResyncStatus persists in-memory resync metadata stats to disk at periodic intervals
+func (p *ReplicationPool) updateResyncStatus(ctx context.Context, objectAPI ObjectLayer) {
 	resyncTimer := time.NewTimer(resyncTimeInterval)
 	defer resyncTimer.Stop()
 
@@ -1980,7 +2001,7 @@ func (p *ReplicationPool) periodicResyncMetaSave(ctx context.Context, objectAPI
 				if updt {
 					brs.LastUpdate = now
 					if err := saveResyncStatus(ctx, bucket, brs, objectAPI); err != nil {
-						logger.LogIf(ctx, fmt.Errorf("Could not save resync metadata to disk for %s - %w", bucket, err))
+						logger.LogIf(ctx, fmt.Errorf("Could not save resync metadata to drive for %s - %w", bucket, err))
 						continue
 					}
 				}
@@ -2041,9 +2062,9 @@ func resyncBucket(ctx context.Context, bucket, arn string, heal bool, objectAPI
 		return
 	}
 
-	// Walk through all object versions - note ascending order of walk needed to ensure delete marker replicated to
-	// target after object version is first created.
-	if err := objectAPI.Walk(ctx, bucket, "", objInfoCh, ObjectOptions{WalkAscending: true}); err != nil {
+	// Walk through all object versions - Walk() is always in ascending order needed to ensure
+	// delete marker replicated to target after object version is first created.
+	if err := objectAPI.Walk(ctx, bucket, "", objInfoCh, ObjectOptions{}); err != nil {
 		logger.LogIf(ctx, err)
 		return
 	}
@@ -2085,13 +2106,15 @@ func resyncBucket(ctx context.Context, bucket, arn string, heal bool, objectAPI
 					DeleteMarkerMTime:     DeleteMarkerMTime{roi.ModTime},
 					DeleteMarker:          roi.DeleteMarker,
 				},
-				Bucket: roi.Bucket,
-				OpType: replication.ExistingObjectReplicationType,
+				Bucket:    roi.Bucket,
+				OpType:    replication.ExistingObjectReplicationType,
+				EventType: ReplicateExistingDelete,
 			}
-			replicateDelete(ctx, doi, objectAPI, ReplicateDelete)
+			replicateDelete(ctx, doi, objectAPI)
 		} else {
 			roi.OpType = replication.ExistingObjectReplicationType
-			replicateObject(ctx, roi, objectAPI, ReplicateExisting)
+			roi.EventType = ReplicateExisting
+			replicateObject(ctx, roi, objectAPI)
 		}
 		_, err = tgt.StatObject(ctx, tgt.Bucket, roi.Name, miniogo.StatObjectOptions{
 			VersionID: roi.VersionID,
@@ -2205,18 +2228,21 @@ func (p *ReplicationPool) loadResync(ctx context.Context, buckets []BucketInfo,
 	for index := range buckets {
 		meta, err := loadBucketResyncMetadata(ctx, buckets[index].Name, objAPI)
 		if err != nil {
-			if errors.Is(err, errVolumeNotFound) {
-				meta = newBucketResyncStatus(buckets[index].Name)
-			} else {
+			if !errors.Is(err, errVolumeNotFound) {
 				logger.LogIf(ctx, err)
-				continue
 			}
+			continue
 		}
+		p.resyncState.Lock()
 		p.resyncState.statusMap[buckets[index].Name] = meta
+		p.resyncState.Unlock()
 	}
 	for index := range buckets {
 		bucket := buckets[index].Name
+		p.resyncState.RLock()
 		m, ok := p.resyncState.statusMap[bucket]
+		p.resyncState.RUnlock()
+
 		if ok {
 			for arn, st := range m.TargetsMap {
 				if st.ResyncStatus == ResyncFailed || st.ResyncStatus == ResyncStarted {
@@ -2283,3 +2309,365 @@ func saveResyncStatus(ctx context.Context, bucket string, brs BucketReplicationR
 	configFile := path.Join(bucketMetaPrefix, bucket, replicationDir, resyncFileName)
 	return saveConfig(ctx, objectAPI, configFile, buf)
 }
+
+// getReplicationDiff returns unreplicated objects in a channel
+func getReplicationDiff(ctx context.Context, objAPI ObjectLayer, bucket string, opts madmin.ReplDiffOpts) (diffCh chan madmin.DiffInfo, err error) {
+	objInfoCh := make(chan ObjectInfo)
+	if err := objAPI.Walk(ctx, bucket, opts.Prefix, objInfoCh, ObjectOptions{}); err != nil {
+		logger.LogIf(ctx, err)
+		return diffCh, err
+	}
+	cfg, err := getReplicationConfig(ctx, bucket)
+	if err != nil {
+		logger.LogIf(ctx, err)
+		return diffCh, err
+	}
+	tgts, err := globalBucketTargetSys.ListBucketTargets(ctx, bucket)
+	if err != nil {
+		logger.LogIf(ctx, err)
+		return diffCh, err
+	}
+	rcfg := replicationConfig{
+		Config:  cfg,
+		remotes: tgts,
+	}
+	diffCh = make(chan madmin.DiffInfo, 4000)
+	go func() {
+		defer close(diffCh)
+		for obj := range objInfoCh {
+			// Ignore object prefixes which are excluded
+			// from versioning via the MinIO bucket versioning extension.
+			if globalBucketVersioningSys.PrefixSuspended(bucket, obj.Name) {
+				continue
+			}
+			roi := getHealReplicateObjectInfo(obj, rcfg)
+			switch roi.ReplicationStatus {
+			case replication.Completed, replication.Replica:
+				if !opts.Verbose {
+					continue
+				}
+				fallthrough
+			default:
+				// ignore pre-existing objects that don't satisfy replication rule(s)
+				if roi.ReplicationStatus.Empty() && !roi.ExistingObjResync.mustResync() {
+					continue
+				}
+				tgtsMap := make(map[string]madmin.TgtDiffInfo)
+				for arn, st := range roi.TargetStatuses {
+					if opts.ARN == "" || opts.ARN == arn {
+						if !opts.Verbose && (st == replication.Completed || st == replication.Replica) {
+							continue
+						}
+						tgtsMap[arn] = madmin.TgtDiffInfo{
+							ReplicationStatus: st.String(),
+						}
+					}
+				}
+				for arn, st := range roi.TargetPurgeStatuses {
+					if opts.ARN == "" || opts.ARN == arn {
+						if !opts.Verbose && st == Complete {
+							continue
+						}
+						t, ok := tgtsMap[arn]
+						if !ok {
+							t = madmin.TgtDiffInfo{}
+						}
+						t.DeleteReplicationStatus = string(st)
+						tgtsMap[arn] = t
+					}
+				}
+				select {
+				case diffCh <- madmin.DiffInfo{
+					Object:                  obj.Name,
+					VersionID:               obj.VersionID,
+					LastModified:            obj.ModTime,
+					IsDeleteMarker:          obj.DeleteMarker,
+					ReplicationStatus:       string(roi.ReplicationStatus),
+					DeleteReplicationStatus: string(roi.VersionPurgeStatus),
+					ReplicationTimestamp:    roi.ReplicationTimestamp,
+					Targets:                 tgtsMap,
+				}:
+				case <-ctx.Done():
+					return
+				}
+			}
+		}
+	}()
+	return diffCh, nil
+}
+
+// QueueReplicationHeal is a wrapper for queueReplicationHeal
+func QueueReplicationHeal(ctx context.Context, bucket string, oi ObjectInfo) {
+	// un-versioned case
+	if oi.VersionID == "" {
+		return
+	}
+	rcfg, _, _ := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
+	tgts, _ := globalBucketTargetSys.ListBucketTargets(ctx, bucket)
+	queueReplicationHeal(ctx, bucket, oi, replicationConfig{
+		Config:  rcfg,
+		remotes: tgts,
+	})
+}
+
+// queueReplicationHeal enqueues objects that failed replication OR eligible for resyncing through
+// an ongoing resync operation or via existing objects replication configuration setting.
+func queueReplicationHeal(ctx context.Context, bucket string, oi ObjectInfo, rcfg replicationConfig) (roi ReplicateObjectInfo) {
+	// un-versioned case
+	if oi.VersionID == "" {
+		return roi
+	}
+
+	if rcfg.Config == nil || rcfg.remotes == nil {
+		return roi
+	}
+	roi = getHealReplicateObjectInfo(oi, rcfg)
+	if !roi.Dsc.ReplicateAny() {
+		return
+	}
+	// early return if replication already done, otherwise we need to determine if this
+	// version is an existing object that needs healing.
+	if oi.ReplicationStatus == replication.Completed && oi.VersionPurgeStatus.Empty() && !roi.ExistingObjResync.mustResync() {
+		return
+	}
+
+	if roi.DeleteMarker || !roi.VersionPurgeStatus.Empty() {
+		versionID := ""
+		dmVersionID := ""
+		if roi.VersionPurgeStatus.Empty() {
+			dmVersionID = roi.VersionID
+		} else {
+			versionID = roi.VersionID
+		}
+
+		dv := DeletedObjectReplicationInfo{
+			DeletedObject: DeletedObject{
+				ObjectName:            roi.Name,
+				DeleteMarkerVersionID: dmVersionID,
+				VersionID:             versionID,
+				ReplicationState:      roi.getReplicationState(roi.Dsc.String(), versionID, true),
+				DeleteMarkerMTime:     DeleteMarkerMTime{roi.ModTime},
+				DeleteMarker:          roi.DeleteMarker,
+			},
+			Bucket:    roi.Bucket,
+			OpType:    replication.HealReplicationType,
+			EventType: ReplicateHealDelete,
+		}
+		// heal delete marker replication failure or versioned delete replication failure
+		if roi.ReplicationStatus == replication.Pending ||
+			roi.ReplicationStatus == replication.Failed ||
+			roi.VersionPurgeStatus == Failed || roi.VersionPurgeStatus == Pending {
+			globalReplicationPool.queueReplicaDeleteTask(dv)
+			return
+		}
+		// if replication status is Complete on DeleteMarker and existing object resync required
+		if roi.ExistingObjResync.mustResync() && (roi.ReplicationStatus == replication.Completed || roi.ReplicationStatus.Empty()) {
+			queueReplicateDeletesWrapper(dv, roi.ExistingObjResync)
+			return
+		}
+		return
+	}
+	if roi.ExistingObjResync.mustResync() {
+		roi.OpType = replication.ExistingObjectReplicationType
+	}
+	switch roi.ReplicationStatus {
+	case replication.Pending, replication.Failed:
+		roi.EventType = ReplicateHeal
+		globalReplicationPool.queueReplicaTask(roi)
+		return
+	}
+	if roi.ExistingObjResync.mustResync() {
+		roi.EventType = ReplicateExisting
+		globalReplicationPool.queueReplicaTask(roi)
+	}
+	return
+}
+
+const mrfTimeInterval = 5 * time.Minute
+
+func (p *ReplicationPool) persistMRF() {
+	var mu sync.Mutex
+	entries := make(map[string]MRFReplicateEntry)
+	mTimer := time.NewTimer(mrfTimeInterval)
+	defer mTimer.Stop()
+	saveMRFToDisk := func(drain bool) {
+		mu.Lock()
+		defer mu.Unlock()
+		if len(entries) == 0 {
+			return
+		}
+		cctx := p.ctx
+		if drain {
+			cctx = context.Background()
+			// drain all mrf entries and save to disk
+			for e := range p.mrfSaveCh {
+				entries[e.versionID] = e
+			}
+		}
+		if err := p.saveMRFEntries(cctx, entries); err != nil {
+			logger.LogOnceIf(p.ctx, fmt.Errorf("Unable to persist replication failures to disk:%w", err), string(replicationSubsystem))
+		}
+		entries = make(map[string]MRFReplicateEntry)
+		return
+	}
+	for {
+		select {
+		case <-mTimer.C:
+			saveMRFToDisk(false)
+			mTimer.Reset(mrfTimeInterval)
+		case <-p.ctx.Done():
+			close(p.mrfSaveCh)
+			saveMRFToDisk(true)
+			return
+		case e, ok := <-p.mrfSaveCh:
+			if !ok {
+				return
+			}
+			var cnt int
+			mu.Lock()
+			entries[e.versionID] = e
+			cnt = len(entries)
+			mu.Unlock()
+			if cnt >= cap(p.mrfSaveCh) || len(p.mrfSaveCh) >= int(0.8*float32(cap(p.mrfSaveCh))) {
+				saveMRFToDisk(true)
+			}
+		}
+	}
+}
+
+func (p *ReplicationPool) queueMRFSave(entry MRFReplicateEntry) {
+	if p == nil {
+		return
+	}
+	select {
+	case <-GlobalContext.Done():
+		return
+	case p.mrfSaveCh <- entry:
+	}
+}
+
+// save mrf entries to mrf_<uuid>.bin
+func (p *ReplicationPool) saveMRFEntries(ctx context.Context, entries map[string]MRFReplicateEntry) error {
+	if len(entries) == 0 {
+		return nil
+	}
+	v := MRFReplicateEntries{
+		Entries: entries,
+		Version: mrfMetaVersionV1,
+	}
+	data := make([]byte, 4, v.Msgsize()+4)
+
+	// Initialize the resync meta header.
+	binary.LittleEndian.PutUint16(data[0:2], resyncMetaFormat)
+	binary.LittleEndian.PutUint16(data[2:4], resyncMetaVersion)
+
+	buf, err := v.MarshalMsg(data)
+	if err != nil {
+		return err
+	}
+
+	configFile := path.Join(replicationMRFDir, mustGetUUID()+".bin")
+	err = saveConfig(ctx, p.objLayer, configFile, buf)
+	return err
+}
+
+// load mrf entries from disk
+func (p *ReplicationPool) loadMRF(fileName string) (re MRFReplicateEntries, e error) {
+	data, err := readConfig(p.ctx, p.objLayer, fileName)
+	if err != nil && err != errConfigNotFound {
+		return re, err
+	}
+	if len(data) == 0 {
+		// Seems to be empty.
+		return re, nil
+	}
+	if len(data) <= 4 {
+		return re, fmt.Errorf("replication mrf: no data")
+	}
+	// Read resync meta header
+	switch binary.LittleEndian.Uint16(data[0:2]) {
+	case mrfMetaFormat:
+	default:
+		return re, fmt.Errorf("replication mrf: unknown format: %d", binary.LittleEndian.Uint16(data[0:2]))
+	}
+	switch binary.LittleEndian.Uint16(data[2:4]) {
+	case mrfMetaVersion:
+	default:
+		return re, fmt.Errorf("replication mrf: unknown version: %d", binary.LittleEndian.Uint16(data[2:4]))
+	}
+	// OK, parse data.
+	if _, err = re.UnmarshalMsg(data[4:]); err != nil {
+		return re, err
+	}
+
+	switch re.Version {
+	case mrfMetaVersionV1:
+	default:
+		return re, fmt.Errorf("unexpected mrf meta version: %d", re.Version)
+	}
+	return re, nil
+}
+
+func (p *ReplicationPool) processMRF() {
+	if p == nil || p.objLayer == nil {
+		return
+	}
+	pTimer := time.NewTimer(mrfTimeInterval)
+	defer pTimer.Stop()
+	for {
+		select {
+		case <-pTimer.C:
+			// skip healing if all targets are offline
+			var offlineCnt int
+			tgts := globalBucketTargetSys.ListTargets(p.ctx, "", "")
+			for _, tgt := range tgts {
+				if globalBucketTargetSys.isOffline(tgt.URL()) {
+					offlineCnt++
+				}
+			}
+			if len(tgts) == offlineCnt {
+				pTimer.Reset(mrfTimeInterval)
+				continue
+			}
+			objCh := make(chan ObjectInfo)
+			cctx, cancelFn := context.WithCancel(p.ctx)
+			if err := p.objLayer.Walk(cctx, minioMetaBucket, replicationMRFDir, objCh, ObjectOptions{}); err != nil {
+				pTimer.Reset(mrfTimeInterval)
+				cancelFn()
+				logger.LogIf(p.ctx, err)
+				continue
+			}
+			for item := range objCh {
+				if err := p.queueMRFHeal(item.Name); err == nil {
+					p.objLayer.DeleteObject(p.ctx, minioMetaBucket, item.Name, ObjectOptions{})
+				}
+			}
+			pTimer.Reset(mrfTimeInterval)
+			cancelFn()
+		case <-p.ctx.Done():
+			return
+		}
+	}
+}
+
+// process sends error logs to the heal channel for an attempt to heal replication.
+func (p *ReplicationPool) queueMRFHeal(file string) error {
+	if p == nil || p.objLayer == nil {
+		return errServerNotInitialized
+	}
+
+	mrfRec, err := p.loadMRF(file)
+	if err != nil {
+		return err
+	}
+	for vID, e := range mrfRec.Entries {
+		oi, err := p.objLayer.GetObjectInfo(p.ctx, e.Bucket, e.Object, ObjectOptions{
+			VersionID: vID,
+		})
+		if err != nil {
+			continue
+		}
+		QueueReplicationHeal(p.ctx, e.Bucket, oi)
+	}
+	return nil
+}

cmd/bucket-stats.go

@@ -26,7 +26,7 @@ import (
 // ReplicationLatency holds information of bucket operations latency, such us uploads
 type ReplicationLatency struct {
 	// Single & Multipart PUTs latency
-	UploadHistogram LastMinuteLatencies
+	UploadHistogram LastMinuteHistogram
 }
 
 // Merge two replication latency into a new one
@@ -41,7 +41,7 @@ func (rl ReplicationLatency) getUploadLatency() (ret map[string]uint64) {
 	avg := rl.UploadHistogram.GetAvgData()
 	for k, v := range avg {
 		// Convert nanoseconds to milliseconds
-		ret[sizeTagToString(k)] = v.avg() / uint64(time.Millisecond)
+		ret[sizeTagToString(k)] = uint64(v.avg() / time.Millisecond)
 	}
 	return
 }

cmd/bucket-targets.go

@@ -19,13 +19,14 @@ package cmd
 
 import (
 	"context"
-	"net/http"
+	"fmt"
+	"net/url"
 	"sync"
 	"time"
 
 	jsoniter "github.com/json-iterator/go"
 	"github.com/minio/madmin-go"
-	minio "github.com/minio/minio-go/v7"
+	"github.com/minio/minio-go/v7"
 	miniogo "github.com/minio/minio-go/v7"
 	"github.com/minio/minio-go/v7/pkg/credentials"
 	"github.com/minio/minio/internal/bucket/replication"
@@ -36,7 +37,9 @@ import (
 )
 
 const (
-	defaultHealthCheckDuration = 30 * time.Second
+	defaultHealthCheckDuration = 5 * time.Second
+	// default interval for reload of all remote target endpoints
+	defaultHealthCheckReloadDuration = 30 * time.Minute
 )
 
 // BucketTargetSys represents bucket targets subsystem
@@ -44,6 +47,113 @@ type BucketTargetSys struct {
 	sync.RWMutex
 	arnRemotesMap map[string]*TargetClient
 	targetsMap    map[string][]madmin.BucketTarget
+	hMutex        sync.RWMutex
+	hc            map[string]epHealth
+	hcClient      *madmin.AnonymousClient
+}
+
+// epHealth struct represents health of a replication target endpoint.
+type epHealth struct {
+	Endpoint string
+	Scheme   string
+	Online   bool
+}
+
+// isOffline returns current liveness result of remote target. Add endpoint to
+// healthcheck map if missing and default to online status
+func (sys *BucketTargetSys) isOffline(ep *url.URL) bool {
+	sys.hMutex.RLock()
+	defer sys.hMutex.RUnlock()
+	if h, ok := sys.hc[ep.Host]; ok {
+		return !h.Online
+	}
+	go sys.initHC(ep)
+	return false
+}
+
+func (sys *BucketTargetSys) initHC(ep *url.URL) {
+	sys.hMutex.Lock()
+	sys.hc[ep.Host] = epHealth{
+		Endpoint: ep.Host,
+		Scheme:   ep.Scheme,
+		Online:   true,
+	}
+	sys.hMutex.Unlock()
+}
+
+// newHCClient initializes an anonymous client for performing health check on the remote endpoints
+func newHCClient() *madmin.AnonymousClient {
+	clnt, e := madmin.NewAnonymousClientNoEndpoint()
+	if e != nil {
+		logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Unable to initialize health check client"), string(replicationSubsystem))
+		return nil
+	}
+	clnt.SetCustomTransport(globalRemoteTargetTransport)
+	return clnt
+}
+
+// heartBeat performs liveness check on remote endpoints.
+func (sys *BucketTargetSys) heartBeat(ctx context.Context) {
+	hcTimer := time.NewTimer(defaultHealthCheckDuration)
+	defer hcTimer.Stop()
+	for {
+		select {
+		case <-hcTimer.C:
+			sys.hMutex.RLock()
+			var eps []madmin.ServerProperties
+			for _, ep := range sys.hc {
+				eps = append(eps, madmin.ServerProperties{Endpoint: ep.Endpoint, Scheme: ep.Scheme})
+			}
+			sys.hMutex.RUnlock()
+
+			if len(eps) > 0 {
+				cctx, cancel := context.WithTimeout(ctx, 30*time.Second)
+				defer cancel()
+				m := map[string]epHealth{}
+				for result := range sys.hcClient.Alive(cctx, madmin.AliveOpts{}, eps...) {
+					var online bool
+					if result.Error == nil {
+						online = result.Online
+					}
+					m[result.Endpoint.Host] = epHealth{
+						Endpoint: result.Endpoint.Host,
+						Scheme:   result.Endpoint.Scheme,
+						Online:   online,
+					}
+				}
+				sys.hMutex.Lock()
+				sys.hc = m
+				sys.hMutex.Unlock()
+			}
+			hcTimer.Reset(defaultHealthCheckDuration)
+		case <-ctx.Done():
+			return
+		}
+	}
+}
+
+// periodically rebuild the healthcheck map from list of targets to clear
+// out stale endpoints
+func (sys *BucketTargetSys) reloadHealthCheckers(ctx context.Context) {
+	m := make(map[string]epHealth)
+	tgts := sys.ListTargets(ctx, "", "")
+	for _, t := range tgts {
+		if _, ok := m[t.Endpoint]; !ok {
+			scheme := "http"
+			if t.Secure {
+				scheme = "https"
+			}
+			m[t.Endpoint] = epHealth{
+				Online:   true,
+				Endpoint: t.Endpoint,
+				Scheme:   scheme,
+			}
+		}
+	}
+	sys.hMutex.Lock()
+	// swap out the map
+	sys.hc = m
+	sys.hMutex.Unlock()
 }
 
 // ListTargets lists bucket targets across tenant or for individual bucket, and returns
@@ -92,9 +202,6 @@ func (sys *BucketTargetSys) Delete(bucket string) {
 		return
 	}
 	for _, t := range tgts {
-		if tgt, ok := sys.arnRemotesMap[t.Arn]; ok && tgt.healthCancelFn != nil {
-			tgt.healthCancelFn()
-		}
 		delete(sys.arnRemotesMap, t.Arn)
 	}
 	delete(sys.targetsMap, bucket)
@@ -117,7 +224,7 @@ func (sys *BucketTargetSys) SetTarget(ctx context.Context, bucket string, tgt *m
 		if minio.ToErrorResponse(err).Code == "NoSuchBucket" {
 			return BucketRemoteTargetNotFound{Bucket: tgt.TargetBucket}
 		}
-		return BucketRemoteConnectionErr{Bucket: tgt.TargetBucket, Err: err}
+		return RemoteTargetConnectionErr{Bucket: tgt.TargetBucket, Err: err}
 	}
 	if tgt.Type == madmin.ReplicationService {
 		if !globalBucketVersioningSys.Enabled(bucket) {
@@ -125,7 +232,7 @@ func (sys *BucketTargetSys) SetTarget(ctx context.Context, bucket string, tgt *m
 		}
 		vcfg, err := clnt.GetBucketVersioning(ctx, tgt.TargetBucket)
 		if err != nil {
-			return BucketRemoteConnectionErr{Bucket: tgt.TargetBucket, Err: err}
+			return RemoteTargetConnectionErr{Bucket: tgt.TargetBucket, Err: err}
 		}
 		if vcfg.Status != string(versioning.Enabled) {
 			return BucketRemoteTargetNotVersioned{Bucket: tgt.TargetBucket}
@@ -153,10 +260,7 @@ func (sys *BucketTargetSys) SetTarget(ctx context.Context, bucket string, tgt *m
 	if !found && !update {
 		newtgts = append(newtgts, *tgt)
 	}
-	// cancel health check for previous target client to avoid leak.
-	if prevClnt, ok := sys.arnRemotesMap[tgt.Arn]; ok && prevClnt.healthCancelFn != nil {
-		prevClnt.healthCancelFn()
-	}
+
 	sys.targetsMap[bucket] = newtgts
 	sys.arnRemotesMap[tgt.Arn] = clnt
 	sys.updateBandwidthLimit(bucket, tgt.BandwidthLimit)
@@ -228,9 +332,6 @@ func (sys *BucketTargetSys) RemoveTarget(ctx context.Context, bucket, arnStr str
 		return BucketRemoteTargetNotFound{Bucket: bucket}
 	}
 	sys.targetsMap[bucket] = targets
-	if tgt, ok := sys.arnRemotesMap[arnStr]; ok && tgt.healthCancelFn != nil {
-		tgt.healthCancelFn()
-	}
 	delete(sys.arnRemotesMap, arnStr)
 	sys.updateBandwidthLimit(bucket, 0)
 	return nil
@@ -259,11 +360,29 @@ func (sys *BucketTargetSys) GetRemoteBucketTargetByArn(ctx context.Context, buck
 }
 
 // NewBucketTargetSys - creates new replication system.
-func NewBucketTargetSys() *BucketTargetSys {
-	return &BucketTargetSys{
+func NewBucketTargetSys(ctx context.Context) *BucketTargetSys {
+	sys := &BucketTargetSys{
 		arnRemotesMap: make(map[string]*TargetClient),
 		targetsMap:    make(map[string][]madmin.BucketTarget),
+		hc:            make(map[string]epHealth),
+		hcClient:      newHCClient(),
 	}
+	// reload healthcheck endpoints map periodically to remove stale endpoints from the map.
+	go func() {
+		rTimer := time.NewTimer(defaultHealthCheckReloadDuration)
+		defer rTimer.Stop()
+		for {
+			select {
+			case <-rTimer.C:
+				sys.reloadHealthCheckers(ctx)
+				rTimer.Reset(defaultHealthCheckReloadDuration)
+			case <-ctx.Done():
+				return
+			}
+		}
+	}()
+	go sys.heartBeat(ctx)
+	return sys
 }
 
 // UpdateAllTargets updates target to reflect metadata updates
@@ -273,17 +392,17 @@ func (sys *BucketTargetSys) UpdateAllTargets(bucket string, tgts *madmin.BucketT
 	}
 	sys.Lock()
 	defer sys.Unlock()
-	if tgts == nil || tgts.Empty() {
-		// remove target and arn association
-		if tgts, ok := sys.targetsMap[bucket]; ok {
-			for _, t := range tgts {
-				if tgt, ok := sys.arnRemotesMap[t.Arn]; ok && tgt.healthCancelFn != nil {
-					tgt.healthCancelFn()
-				}
-				delete(sys.arnRemotesMap, t.Arn)
-			}
+
+	// Remove existingtarget and arn association
+	if tgts, ok := sys.targetsMap[bucket]; ok {
+		for _, t := range tgts {
+			delete(sys.arnRemotesMap, t.Arn)
 		}
 		delete(sys.targetsMap, bucket)
+	}
+
+	// No need for more if not adding anything
+	if tgts == nil || tgts.Empty() {
 		sys.updateBandwidthLimit(bucket, 0)
 		return
 	}
@@ -325,25 +444,16 @@ func (sys *BucketTargetSys) set(bucket BucketInfo, meta BucketMetadata) {
 	sys.targetsMap[bucket.Name] = cfg.Targets
 }
 
-// getRemoteTargetInstanceTransport contains a singleton roundtripper.
-var (
-	getRemoteTargetInstanceTransport     http.RoundTripper
-	getRemoteTargetInstanceTransportOnce sync.Once
-)
-
 // Returns a minio-go Client configured to access remote host described in replication target config.
 func (sys *BucketTargetSys) getRemoteTargetClient(tcfg *madmin.BucketTarget) (*TargetClient, error) {
 	config := tcfg.Credentials
 	creds := credentials.NewStaticV4(config.AccessKey, config.SecretKey, "")
-	getRemoteTargetInstanceTransportOnce.Do(func() {
-		getRemoteTargetInstanceTransport = NewRemoteTargetHTTPTransport()
-	})
 
 	api, err := minio.New(tcfg.Endpoint, &miniogo.Options{
 		Creds:     creds,
 		Secure:    tcfg.Secure,
 		Region:    tcfg.Region,
-		Transport: getRemoteTargetInstanceTransport,
+		Transport: globalRemoteTargetTransport,
 	})
 	if err != nil {
 		return nil, err
@@ -352,10 +462,6 @@ func (sys *BucketTargetSys) getRemoteTargetClient(tcfg *madmin.BucketTarget) (*T
 	if tcfg.HealthCheckDuration >= 1 { // require minimum health check duration of 1 sec.
 		hcDuration = tcfg.HealthCheckDuration
 	}
-	cancelFn, err := api.HealthCheck(hcDuration)
-	if err != nil {
-		return nil, err
-	}
 	tc := &TargetClient{
 		Client:              api,
 		healthCheckDuration: hcDuration,
@@ -363,9 +469,10 @@ func (sys *BucketTargetSys) getRemoteTargetClient(tcfg *madmin.BucketTarget) (*T
 		Bucket:              tcfg.TargetBucket,
 		StorageClass:        tcfg.StorageClass,
 		disableProxy:        tcfg.DisableProxy,
-		healthCancelFn:      cancelFn,
 		ARN:                 tcfg.Arn,
 		ResetID:             tcfg.ResetID,
+		Endpoint:            tcfg.Endpoint,
+		Secure:              tcfg.Secure,
 	}
 	return tc, nil
 }
@@ -439,7 +546,8 @@ type TargetClient struct {
 	replicateSync       bool
 	StorageClass        string // storage class on remote
 	disableProxy        bool
-	healthCancelFn      context.CancelFunc // cancellation function for client healthcheck
-	ARN                 string             // ARN to uniquely identify remote target
+	ARN                 string // ARN to uniquely identify remote target
 	ResetID             string
+	Endpoint            string
+	Secure              bool
 }

cmd/bucket-versioning-handler.go

@@ -97,7 +97,8 @@ func (api objectAPIHandlers) PutBucketVersioningHandler(w http.ResponseWriter, r
 		return
 	}
 
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketVersioningConfig, configData); err != nil {
+	updatedAt, err := globalBucketMetadataSys.Update(ctx, bucket, bucketVersioningConfig, configData)
+	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -111,6 +112,7 @@ func (api objectAPIHandlers) PutBucketVersioningHandler(w http.ResponseWriter, r
 		Type:       madmin.SRBucketMetaTypeVersionConfig,
 		Bucket:     bucket,
 		Versioning: &cfgStr,
+		UpdatedAt:  updatedAt,
 	}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -141,7 +143,7 @@ func (api objectAPIHandlers) GetBucketVersioningHandler(w http.ResponseWriter, r
 	}
 
 	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}

cmd/build-constants.go

@@ -1,4 +1,4 @@
-// Copyright (c) 2015-2021 MinIO, Inc.
+// Copyright (c) 2015-2022 MinIO, Inc.
 //
 // This file is part of MinIO Object Storage stack
 //
@@ -37,4 +37,7 @@ var (
 
 	// ShortCommitID - first 12 characters from CommitID.
 	ShortCommitID = "DEVELOPMENT.GOGET"
+
+	// CopyrightYear - dynamic value of the copyright end year
+	CopyrightYear = "0000"
 )

cmd/callhome.go

@@ -134,3 +134,17 @@ func performCallhome(ctx context.Context) {
 		logger.LogIf(ctx, fmt.Errorf("Unable to perform callhome: %w", err))
 	}
 }
+
+const (
+	callhomeURL    = "https://subnet.min.io/api/callhome"
+	callhomeURLDev = "http://localhost:9000/api/callhome"
+)
+
+func sendCallhomeInfo(ch CallhomeInfo) error {
+	url := callhomeURL
+	if globalIsCICD {
+		url = callhomeURLDev
+	}
+	_, err := globalSubnetConfig.Post(url, ch)
+	return err
+}

cmd/common-main.go

@@ -48,6 +48,7 @@ import (
 	"github.com/inconshreveable/mousetrap"
 	dns2 "github.com/miekg/dns"
 	"github.com/minio/cli"
+	consoleoauth2 "github.com/minio/console/pkg/auth/idp/oauth2"
 	consoleCerts "github.com/minio/console/pkg/certs"
 	"github.com/minio/console/restapi"
 	"github.com/minio/console/restapi/operations"
@@ -102,6 +103,11 @@ func init() {
 		PersistOnFailure: false,
 	}
 
+	t, _ := minioVersionToReleaseTime(Version)
+	if !t.IsZero() {
+		globalVersionUnix = uint64(t.Unix())
+	}
+
 	globalIsCICD = env.Get("MINIO_CI_CD", "") != "" || env.Get("CI", "") != ""
 
 	containers := IsKubernetes() || IsDocker() || IsBOSH() || IsDCOS() || IsPCFTile()
@@ -200,28 +206,6 @@ func minioConfigToConsoleFeatures() {
 	if globalLDAPConfig.Enabled {
 		os.Setenv("CONSOLE_LDAP_ENABLED", config.EnableOn)
 	}
-	// if IDP is enabled, set IDP environment variables
-	if globalOpenIDConfig.ProviderCfgs[config.Default] != nil {
-		os.Setenv("CONSOLE_IDP_URL", globalOpenIDConfig.ProviderCfgs[config.Default].URL.String())
-		os.Setenv("CONSOLE_IDP_CLIENT_ID", globalOpenIDConfig.ProviderCfgs[config.Default].ClientID)
-		os.Setenv("CONSOLE_IDP_SECRET", globalOpenIDConfig.ProviderCfgs[config.Default].ClientSecret)
-		os.Setenv("CONSOLE_IDP_HMAC_SALT", globalDeploymentID)
-		os.Setenv("CONSOLE_IDP_HMAC_PASSPHRASE", globalOpenIDConfig.ProviderCfgs[config.Default].ClientID)
-		os.Setenv("CONSOLE_IDP_SCOPES", strings.Join(globalOpenIDConfig.ProviderCfgs[config.Default].DiscoveryDoc.ScopesSupported, ","))
-		if globalOpenIDConfig.ProviderCfgs[config.Default].ClaimUserinfo {
-			os.Setenv("CONSOLE_IDP_USERINFO", config.EnableOn)
-		}
-		if globalOpenIDConfig.ProviderCfgs[config.Default].RedirectURIDynamic {
-			// Enable dynamic redirect-uri's based on incoming 'host' header,
-			// Overrides any other callback URL.
-			os.Setenv("CONSOLE_IDP_CALLBACK_DYNAMIC", config.EnableOn)
-		}
-		if globalOpenIDConfig.ProviderCfgs[config.Default].RedirectURI != "" {
-			os.Setenv("CONSOLE_IDP_CALLBACK", globalOpenIDConfig.ProviderCfgs[config.Default].RedirectURI)
-		} else {
-			os.Setenv("CONSOLE_IDP_CALLBACK", getConsoleEndpoints()[0]+"/oauth_callback")
-		}
-	}
 	os.Setenv("CONSOLE_MINIO_REGION", globalSite.Region)
 	os.Setenv("CONSOLE_CERT_PASSWD", env.Get("MINIO_CERT_PASSWD", ""))
 	if globalSubnetConfig.License != "" {
@@ -235,6 +219,29 @@ func minioConfigToConsoleFeatures() {
 	}
 }
 
+func buildOpenIDConsoleConfig() consoleoauth2.OpenIDPCfg {
+	m := make(map[string]consoleoauth2.ProviderConfig, len(globalOpenIDConfig.ProviderCfgs))
+	for name, cfg := range globalOpenIDConfig.ProviderCfgs {
+		callback := getConsoleEndpoints()[0] + "/oauth_callback"
+		if cfg.RedirectURI != "" {
+			callback = cfg.RedirectURI
+		}
+		m[name] = consoleoauth2.ProviderConfig{
+			URL:                     cfg.URL.String(),
+			DisplayName:             cfg.DisplayName,
+			ClientID:                cfg.ClientID,
+			ClientSecret:            cfg.ClientSecret,
+			HMACSalt:                globalDeploymentID,
+			HMACPassphrase:          cfg.ClientID,
+			Scopes:                  strings.Join(cfg.DiscoveryDoc.ScopesSupported, ","),
+			Userinfo:                cfg.ClaimUserinfo,
+			RedirectCallbackDynamic: cfg.RedirectURIDynamic,
+			RedirectCallback:        callback,
+		}
+	}
+	return m
+}
+
 func initConsoleServer() (*restapi.Server, error) {
 	// unset all console_ environment variables.
 	for _, cenv := range env.List(consolePrefix) {
@@ -268,6 +275,12 @@ func initConsoleServer() (*restapi.Server, error) {
 		api.Logger = noLog
 	}
 
+	// Pass in console application config. This needs to happen before the
+	// ConfigureAPI() call.
+	restapi.GlobalMinIOConfig = restapi.MinIOConfig{
+		OpenIDProviders: buildOpenIDConsoleConfig(),
+	}
+
 	server := restapi.NewServer(api)
 	// register all APIs
 	server.ConfigureAPI()
@@ -425,7 +438,6 @@ func handleCommonCmdArgs(ctx *cli.Context) {
 		if err != nil {
 			logger.FatalIf(err, "Unable to get free port for console on the host")
 		}
-		globalMinioConsolePortAuto = true
 		consoleAddr = net.JoinHostPort("", p.String())
 	}
 
@@ -490,8 +502,13 @@ func parsEnvEntry(envEntry string) (envKV, error) {
 			Skip: true,
 		}, nil
 	}
-	const envSeparator = "="
-	envTokens := strings.SplitN(strings.TrimSpace(strings.TrimPrefix(envEntry, "export")), envSeparator, 2)
+	if strings.HasPrefix(envEntry, "#") {
+		// Skip commented lines
+		return envKV{
+			Skip: true,
+		}, nil
+	}
+	envTokens := strings.SplitN(strings.TrimSpace(strings.TrimPrefix(envEntry, "export")), config.EnvSeparator, 2)
 	if len(envTokens) != 2 {
 		return envKV{}, fmt.Errorf("envEntry malformed; %s, expected to be of form 'KEY=value'", envEntry)
 	}
@@ -499,13 +516,6 @@ func parsEnvEntry(envEntry string) (envKV, error) {
 	key := envTokens[0]
 	val := envTokens[1]
 
-	if strings.HasPrefix(key, "#") {
-		// Skip commented lines
-		return envKV{
-			Skip: true,
-		}, nil
-	}
-
 	// Remove quotes from the value if found
 	if len(val) >= 2 {
 		quote := val[0]
@@ -524,11 +534,8 @@ func parsEnvEntry(envEntry string) (envKV, error) {
 // the environment values from a file, in the form "key, value".
 // in a structured form.
 func minioEnvironFromFile(envConfigFile string) ([]envKV, error) {
-	f, err := os.Open(envConfigFile)
+	f, err := Open(envConfigFile)
 	if err != nil {
-		if os.IsNotExist(err) { // ignore if file doesn't exist.
-			return nil, nil
-		}
 		return nil, err
 	}
 	defer f.Close()
@@ -624,7 +631,7 @@ func loadEnvVarsFromFiles() {
 
 	if env.IsSet(config.EnvConfigEnvFile) {
 		ekvs, err := minioEnvironFromFile(env.Get(config.EnvConfigEnvFile, ""))
-		if err != nil {
+		if err != nil && !os.IsNotExist(err) {
 			logger.Fatal(err, "Unable to read the config environment file")
 		}
 		for _, ekv := range ekvs {
@@ -787,17 +794,22 @@ func handleCommonEnvVars() {
 		}
 		globalActiveCred = cred
 	}
+}
 
+// Initialize KMS global variable after valiadating and loading the configuration.
+// It depends on KMS env variables and global cli flags.
+func handleKMSConfig() {
 	switch {
 	case env.IsSet(config.EnvKMSSecretKey) && env.IsSet(config.EnvKESEndpoint):
 		logger.Fatal(errors.New("ambigious KMS configuration"), fmt.Sprintf("The environment contains %q as well as %q", config.EnvKMSSecretKey, config.EnvKESEndpoint))
 	}
 
 	if env.IsSet(config.EnvKMSSecretKey) {
-		GlobalKMS, err = kms.Parse(env.Get(config.EnvKMSSecretKey, ""))
+		KMS, err := kms.Parse(env.Get(config.EnvKMSSecretKey, ""))
 		if err != nil {
 			logger.Fatal(err, "Unable to parse the KMS secret key inherited from the shell environment")
 		}
+		GlobalKMS = KMS
 	}
 	if env.IsSet(config.EnvKESEndpoint) {
 		var endpoints []string
@@ -817,43 +829,56 @@ func handleCommonEnvVars() {
 				endpoints = append(endpoints, strings.Join(lbls, ""))
 			}
 		}
-		// Manually load the certificate and private key into memory.
-		// We need to check whether the private key is encrypted, and
-		// if so, decrypt it using the user-provided password.
-		certBytes, err := os.ReadFile(env.Get(config.EnvKESClientCert, ""))
-		if err != nil {
-			logger.Fatal(err, "Unable to load KES client certificate as specified by the shell environment")
-		}
-		keyBytes, err := os.ReadFile(env.Get(config.EnvKESClientKey, ""))
-		if err != nil {
-			logger.Fatal(err, "Unable to load KES client private key as specified by the shell environment")
-		}
-		privateKeyPEM, rest := pem.Decode(bytes.TrimSpace(keyBytes))
-		if len(rest) != 0 {
-			logger.Fatal(errors.New("private key contains additional data"), "Unable to load KES client private key as specified by the shell environment")
-		}
-		if x509.IsEncryptedPEMBlock(privateKeyPEM) {
-			keyBytes, err = x509.DecryptPEMBlock(privateKeyPEM, []byte(env.Get(config.EnvKESClientPassword, "")))
-			if err != nil {
-				logger.Fatal(err, "Unable to decrypt KES client private key as specified by the shell environment")
-			}
-			keyBytes = pem.EncodeToMemory(&pem.Block{Type: privateKeyPEM.Type, Bytes: keyBytes})
-		}
-		certificate, err := tls.X509KeyPair(certBytes, keyBytes)
-		if err != nil {
-			logger.Fatal(err, "Unable to load KES client certificate as specified by the shell environment")
-		}
 		rootCAs, err := certs.GetRootCAs(env.Get(config.EnvKESServerCA, globalCertsCADir.Get()))
 		if err != nil {
 			logger.Fatal(err, fmt.Sprintf("Unable to load X.509 root CAs for KES from %q", env.Get(config.EnvKESServerCA, globalCertsCADir.Get())))
 		}
 
+		loadX509KeyPair := func(certFile, keyFile string) (tls.Certificate, error) {
+			// Manually load the certificate and private key into memory.
+			// We need to check whether the private key is encrypted, and
+			// if so, decrypt it using the user-provided password.
+			certBytes, err := os.ReadFile(certFile)
+			if err != nil {
+				return tls.Certificate{}, fmt.Errorf("Unable to load KES client certificate as specified by the shell environment: %v", err)
+			}
+			keyBytes, err := os.ReadFile(keyFile)
+			if err != nil {
+				return tls.Certificate{}, fmt.Errorf("Unable to load KES client private key as specified by the shell environment: %v", err)
+			}
+			privateKeyPEM, rest := pem.Decode(bytes.TrimSpace(keyBytes))
+			if len(rest) != 0 {
+				return tls.Certificate{}, errors.New("Unable to load KES client private key as specified by the shell environment: private key contains additional data")
+			}
+			if x509.IsEncryptedPEMBlock(privateKeyPEM) {
+				keyBytes, err = x509.DecryptPEMBlock(privateKeyPEM, []byte(env.Get(config.EnvKESClientPassword, "")))
+				if err != nil {
+					return tls.Certificate{}, fmt.Errorf("Unable to decrypt KES client private key as specified by the shell environment: %v", err)
+				}
+				keyBytes = pem.EncodeToMemory(&pem.Block{Type: privateKeyPEM.Type, Bytes: keyBytes})
+			}
+			certificate, err := tls.X509KeyPair(certBytes, keyBytes)
+			if err != nil {
+				return tls.Certificate{}, fmt.Errorf("Unable to load KES client certificate as specified by the shell environment: %v", err)
+			}
+			return certificate, nil
+		}
+
+		reloadCertEvents := make(chan tls.Certificate, 1)
+		certificate, err := certs.NewCertificate(env.Get(config.EnvKESClientCert, ""), env.Get(config.EnvKESClientKey, ""), loadX509KeyPair)
+		if err != nil {
+			logger.Fatal(err, "Failed to load KES client certificate")
+		}
+		certificate.Watch(context.Background(), 15*time.Minute, syscall.SIGHUP)
+		certificate.Notify(reloadCertEvents)
+
 		defaultKeyID := env.Get(config.EnvKESKeyName, "")
 		KMS, err := kms.NewWithConfig(kms.Config{
-			Endpoints:    endpoints,
-			DefaultKeyID: defaultKeyID,
-			Certificate:  certificate,
-			RootCAs:      rootCAs,
+			Endpoints:        endpoints,
+			DefaultKeyID:     defaultKeyID,
+			Certificate:      certificate,
+			ReloadCertEvents: reloadCertEvents,
+			RootCAs:          rootCAs,
 		})
 		if err != nil {
 			logger.Fatal(err, "Unable to initialize a connection to KES as specified by the shell environment")
@@ -863,7 +888,7 @@ func handleCommonEnvVars() {
 		// This implicitly checks that we can communicate to KES. We don't treat
 		// a policy error as failure condition since MinIO may not have the permission
 		// to create keys - just to generate/decrypt data encryption keys.
-		if err = KMS.CreateKey(defaultKeyID); err != nil && !errors.Is(err, kes.ErrKeyExists) && !errors.Is(err, kes.ErrNotAllowed) {
+		if err = KMS.CreateKey(context.Background(), defaultKeyID); err != nil && !errors.Is(err, kes.ErrKeyExists) && !errors.Is(err, kes.ErrNotAllowed) {
 			logger.Fatal(err, "Unable to initialize a connection to KES as specified by the shell environment")
 		}
 		GlobalKMS = KMS
@@ -903,7 +928,7 @@ func getTLSConfig() (x509Certs []*x509.Certificate, manager *certs.Manager, secu
 	// Therefore, we read all filenames in the cert directory and check
 	// for each directory whether it contains a public.crt and private.key.
 	// If so, we try to add it to certificate manager.
-	root, err := os.Open(globalCertsDir.Get())
+	root, err := Open(globalCertsDir.Get())
 	if err != nil {
 		return nil, nil, false, err
 	}
@@ -922,7 +947,7 @@ func getTLSConfig() (x509Certs []*x509.Certificate, manager *certs.Manager, secu
 			continue
 		}
 		if file.Mode()&os.ModeSymlink == os.ModeSymlink {
-			file, err = os.Stat(filepath.Join(root.Name(), file.Name()))
+			file, err = Stat(filepath.Join(root.Name(), file.Name()))
 			if err != nil {
 				// not accessible ignore
 				continue

cmd/common-main_test.go

@@ -136,6 +136,7 @@ export MINIO_ROOT_PASSWORD=minio123`,
 		},
 		{
 			`
+# simple comment
 # MINIO_ROOT_USER=minioadmin
 # MINIO_ROOT_PASSWORD=minioadmin
 MINIO_ROOT_USER=minio

cmd/config-current.go

@@ -338,7 +338,7 @@ func validateSubSysConfig(s config.Config, subSys string, objAPI ObjectLayer) er
 			etcdClnt.Close()
 		}
 	case config.IdentityOpenIDSubSys:
-		if _, err := openid.LookupConfig(s[config.IdentityOpenIDSubSys],
+		if _, err := openid.LookupConfig(s,
 			NewGatewayHTTPTransport(), xhttp.DrainBody, globalSite.Region); err != nil {
 			return err
 		}
@@ -364,7 +364,7 @@ func validateSubSysConfig(s config.Config, subSys string, objAPI ObjectLayer) er
 			return err
 		}
 	case config.SubnetSubSys:
-		if _, err := subnet.LookupConfig(s[config.SubnetSubSys][config.Default]); err != nil {
+		if _, err := subnet.LookupConfig(s[config.SubnetSubSys][config.Default], nil); err != nil {
 			return err
 		}
 	case config.CallhomeSubSys:
@@ -471,14 +471,12 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 	}
 
 	if etcdCfg.Enabled {
-		if globalEtcdClient == nil {
-			globalEtcdClient, err = etcd.New(etcdCfg)
-			if err != nil {
-				if globalIsGateway {
-					logger.FatalIf(err, "Unable to initialize etcd config")
-				} else {
-					logger.LogIf(ctx, fmt.Errorf("Unable to initialize etcd config: %w", err))
-				}
+		globalEtcdClient, err = etcd.New(etcdCfg)
+		if err != nil {
+			if globalIsGateway {
+				logger.FatalIf(err, "Unable to initialize etcd config")
+			} else {
+				logger.LogIf(ctx, fmt.Errorf("Unable to initialize etcd config: %w", err))
 			}
 		}
 
@@ -486,7 +484,7 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 			if globalDNSConfig != nil {
 				// if global DNS is already configured, indicate with a warning, incase
 				// users are confused.
-				logger.LogIf(ctx, fmt.Errorf("DNS store is already configured with %s, not using etcd for DNS store", globalDNSConfig))
+				logger.LogIf(ctx, fmt.Errorf("DNS store is already configured with %s, etcd is not used for DNS store", globalDNSConfig))
 			} else {
 				globalDNSConfig, err = dns.NewCoreDNS(etcdCfg.Config,
 					dns.DomainNames(globalDomainNames),
@@ -518,16 +516,6 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 		logger.LogIf(ctx, fmt.Errorf("Invalid site configuration: %w", err))
 	}
 
-	apiConfig, err := api.LookupConfig(s[config.APISubSys][config.Default])
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Invalid api configuration: %w", err))
-	}
-
-	// Initialize remote instance transport once.
-	getRemoteInstanceTransportOnce.Do(func() {
-		getRemoteInstanceTransport = newGatewayHTTPTransport(apiConfig.RemoteTransportDeadline)
-	})
-
 	globalCacheConfig, err = cache.LookupConfig(s[config.CacheSubSys][config.Default])
 	if err != nil {
 		if globalIsGateway {
@@ -560,56 +548,19 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 		logger.LogIf(ctx, fmt.Errorf("CRITICAL: enabling %s is not recommended in a production environment", xtls.EnvIdentityTLSSkipVerify))
 	}
 
-	globalOpenIDConfig, err = openid.LookupConfig(s[config.IdentityOpenIDSubSys],
-		NewGatewayHTTPTransport(), xhttp.DrainBody, globalSite.Region)
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Unable to initialize OpenID: %w", err))
-	}
-
-	globalLDAPConfig, err = xldap.Lookup(s[config.IdentityLDAPSubSys][config.Default],
-		globalRootCAs)
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Unable to parse LDAP configuration: %w", err))
-	}
-
-	authNPluginCfg, err := idplugin.LookupConfig(s[config.IdentityPluginSubSys][config.Default],
-		NewGatewayHTTPTransport(), xhttp.DrainBody, globalSite.Region)
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Unable to initialize AuthNPlugin: %w", err))
-	}
-	globalAuthNPlugin = idplugin.New(authNPluginCfg)
-
-	authZPluginCfg, err := polplugin.LookupConfig(s[config.PolicyPluginSubSys][config.Default],
-		NewGatewayHTTPTransport(), xhttp.DrainBody)
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Unable to initialize AuthZPlugin: %w", err))
-	}
-	if authZPluginCfg.URL == nil {
-		opaCfg, err := opa.LookupConfig(s[config.PolicyOPASubSys][config.Default],
-			NewGatewayHTTPTransport(), xhttp.DrainBody)
-		if err != nil {
-			logger.LogIf(ctx, fmt.Errorf("Unable to initialize AuthZPlugin from legacy OPA config: %w", err))
-		} else {
-			authZPluginCfg.URL = opaCfg.URL
-			authZPluginCfg.AuthToken = opaCfg.AuthToken
-			authZPluginCfg.Transport = opaCfg.Transport
-			authZPluginCfg.CloseRespFn = opaCfg.CloseRespFn
-		}
-	}
-
-	setGlobalAuthZPlugin(polplugin.New(authZPluginCfg))
-
-	globalSubnetConfig, err = subnet.LookupConfig(s[config.SubnetSubSys][config.Default])
+	globalSubnetConfig, err = subnet.LookupConfig(s[config.SubnetSubSys][config.Default], globalProxyTransport)
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("Unable to parse subnet configuration: %w", err))
 	}
 
-	globalConfigTargetList, err = notify.GetNotificationTargets(GlobalContext, s, NewGatewayHTTPTransport(), false)
+	transport := NewGatewayHTTPTransport()
+
+	globalConfigTargetList, err = notify.GetNotificationTargets(GlobalContext, s, transport, false)
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("Unable to initialize notification target(s): %w", err))
 	}
 
-	globalEnvTargetList, err = notify.GetNotificationTargets(GlobalContext, newServerConfig(), NewGatewayHTTPTransport(), true)
+	globalEnvTargetList, err = notify.GetNotificationTargets(GlobalContext, newServerConfig(), transport, true)
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("Unable to initialize notification target(s): %w", err))
 	}
@@ -636,6 +587,11 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 			setDriveCounts = objAPI.SetDriveCounts()
 		}
 		globalAPIConfig.init(apiConfig, setDriveCounts)
+
+		// Initialize remote instance transport once.
+		getRemoteInstanceTransportOnce.Do(func() {
+			getRemoteInstanceTransport = newGatewayHTTPTransport(apiConfig.RemoteTransportDeadline)
+		})
 	case config.CompressionSubSys:
 		cmpCfg, err := compress.LookupConfig(s[config.CompressionSubSys][config.Default])
 		if err != nil {
@@ -678,8 +634,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 				loggerCfg.HTTP[n] = l
 			}
 		}
-		err = logger.UpdateSystemTargets(loggerCfg)
-		if err != nil {
+		if err = logger.UpdateSystemTargets(loggerCfg); err != nil {
 			logger.LogIf(ctx, fmt.Errorf("Unable to update logger webhook config: %w", err))
 		}
 	case config.AuditWebhookSubSys:
@@ -697,8 +652,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 			}
 		}
 
-		err = logger.UpdateAuditWebhookTargets(loggerCfg)
-		if err != nil {
+		if err = logger.UpdateAuditWebhookTargets(loggerCfg); err != nil {
 			logger.LogIf(ctx, fmt.Errorf("Unable to update audit webhook targets: %w", err))
 		}
 	case config.AuditKafkaSubSys:
@@ -712,8 +666,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 				loggerCfg.AuditKafka[n] = l
 			}
 		}
-		err = logger.UpdateAuditKafkaTargets(loggerCfg)
-		if err != nil {
+		if err = logger.UpdateAuditKafkaTargets(loggerCfg); err != nil {
 			logger.LogIf(ctx, fmt.Errorf("Unable to update audit kafka targets: %w", err))
 		}
 	case config.StorageClassSubSys:
@@ -852,7 +805,7 @@ func newSrvConfig(objAPI ObjectLayer) error {
 	globalServerConfigMu.Unlock()
 
 	// Save config into file.
-	return saveServerConfig(GlobalContext, objAPI, globalServerConfig)
+	return saveServerConfig(GlobalContext, objAPI, srvCfg)
 }
 
 func getValidConfig(objAPI ObjectLayer) (config.Config, error) {

cmd/config-encrypted.go

@@ -69,8 +69,14 @@ func migrateIAMConfigsEtcdToEncrypted(ctx context.Context, client *etcd.Client)
 		return err
 	}
 
+	// If backend doesn't have this file means we have already
+	// attempted then migration
+	if !encrypted {
+		return nil
+	}
+
 	if encrypted && GlobalKMS != nil {
-		stat, err := GlobalKMS.Stat()
+		stat, err := GlobalKMS.Stat(ctx)
 		if err != nil {
 			return err
 		}
@@ -109,6 +115,8 @@ func migrateIAMConfigsEtcdToEncrypted(ctx context.Context, client *etcd.Client)
 							return fmt.Errorf("Decrypting IAM config failed %w, possibly credentials are incorrect", err)
 						}
 					}
+				} else {
+					return fmt.Errorf("Decrypting IAM config failed %w, possibly credentials are incorrect", err)
 				}
 			}
 			data = pdata
@@ -131,6 +139,7 @@ func migrateIAMConfigsEtcdToEncrypted(ctx context.Context, client *etcd.Client)
 	if encrypted && GlobalKMS != nil {
 		logger.Info("Migration of encrypted IAM config data completed. All data is now encrypted on etcd.")
 	}
+
 	return deleteKeyEtcd(ctx, client, backendEncryptedFile)
 }
 
@@ -139,51 +148,63 @@ func migrateConfigPrefixToEncrypted(objAPI ObjectLayer, encrypted bool) error {
 		return nil
 	}
 	if encrypted && GlobalKMS != nil {
-		stat, err := GlobalKMS.Stat()
+		stat, err := GlobalKMS.Stat(context.Background())
 		if err != nil {
 			return err
 		}
 		logger.Info(fmt.Sprintf("Attempting to re-encrypt config, IAM users and policies on MinIO with %q (%s)", stat.DefaultKey, stat.Name))
 	}
 
-	var marker string
-	for {
-		res, err := objAPI.ListObjects(GlobalContext, minioMetaBucket, minioConfigPrefix, marker, "", maxObjectList)
+	results := make(chan ObjectInfo)
+	if err := objAPI.Walk(GlobalContext, minioMetaBucket, minioConfigPrefix, results, ObjectOptions{}); err != nil {
+		return err
+	}
+
+	for obj := range results {
+		data, err := readConfig(GlobalContext, objAPI, obj.Name)
 		if err != nil {
 			return err
 		}
-		for _, obj := range res.Objects {
-			data, err := readConfig(GlobalContext, objAPI, obj.Name)
+
+		if !utf8.Valid(data) {
+			pdata, err := madmin.DecryptData(globalActiveCred.String(), bytes.NewReader(data))
+			if err != nil {
+				if GlobalKMS != nil {
+					pdata, err = config.DecryptBytes(GlobalKMS, data, kms.Context{
+						minioMetaBucket: path.Join(minioMetaBucket, obj.Name),
+					})
+					if err != nil {
+						pdata, err = config.DecryptBytes(GlobalKMS, data, kms.Context{
+							minioMetaBucket: obj.Name,
+						})
+						if err != nil {
+							return fmt.Errorf("Decrypting IAM config failed %w, possibly credentials are incorrect", err)
+						}
+					}
+				} else {
+					return fmt.Errorf("Decrypting IAM config failed %w, possibly credentials are incorrect", err)
+				}
+			}
+			data = pdata
+		}
+
+		if GlobalKMS != nil {
+			data, err = config.EncryptBytes(GlobalKMS, data, kms.Context{
+				obj.Bucket: path.Join(obj.Bucket, obj.Name),
+			})
 			if err != nil {
 				return err
 			}
-
-			if !utf8.Valid(data) {
-				data, err = madmin.DecryptData(globalActiveCred.String(), bytes.NewReader(data))
-				if err != nil {
-					return fmt.Errorf("Decrypting config failed %w, possibly credentials are incorrect", err)
-				}
-			}
-			if GlobalKMS != nil {
-				data, err = config.EncryptBytes(GlobalKMS, data, kms.Context{
-					obj.Bucket: path.Join(obj.Bucket, obj.Name),
-				})
-				if err != nil {
-					return err
-				}
-			}
-			if err = saveConfig(GlobalContext, objAPI, obj.Name, data); err != nil {
-				return err
-			}
 		}
 
-		if !res.IsTruncated {
-			break
+		if err = saveConfig(GlobalContext, objAPI, obj.Name, data); err != nil {
+			return err
 		}
-		marker = res.NextMarker
 	}
+
 	if encrypted && GlobalKMS != nil {
 		logger.Info("Migration of encrypted config data completed. All config data is now encrypted.")
 	}
+
 	return deleteConfig(GlobalContext, objAPI, backendEncryptedFile)
 }

cmd/config-migrate_test.go

@@ -37,11 +37,7 @@ func TestServerConfigMigrateV1(t *testing.T) {
 	if err != nil {
 		t.Fatalf("Init Test config failed")
 	}
-	rootPath, err := ioutil.TempDir(globalTestTmpDir, "minio-")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(rootPath)
+	rootPath := t.TempDir()
 	globalConfigDir = &ConfigDir{path: rootPath}
 
 	globalObjLayerMutex.Lock()
@@ -74,13 +70,7 @@ func TestServerConfigMigrateV1(t *testing.T) {
 // Test if all migrate code returns nil when config file does not
 // exist
 func TestServerConfigMigrateInexistentConfig(t *testing.T) {
-	rootPath, err := ioutil.TempDir(globalTestTmpDir, "minio-")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(rootPath)
-
-	globalConfigDir = &ConfigDir{path: rootPath}
+	globalConfigDir = &ConfigDir{path: t.TempDir()}
 
 	if err := migrateV2ToV3(); err != nil {
 		t.Fatal("migrate v2 to v3 should succeed when no config file is found")
@@ -164,11 +154,7 @@ func TestServerConfigMigrateInexistentConfig(t *testing.T) {
 
 // Test if a config migration from v2 to v33 is successfully done
 func TestServerConfigMigrateV2toV33(t *testing.T) {
-	rootPath, err := ioutil.TempDir(globalTestTmpDir, "minio-")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(rootPath)
+	rootPath := t.TempDir()
 
 	globalConfigDir = &ConfigDir{path: rootPath}
 
@@ -234,11 +220,7 @@ func TestServerConfigMigrateV2toV33(t *testing.T) {
 
 // Test if all migrate code returns error with corrupted config files
 func TestServerConfigMigrateFaultyConfig(t *testing.T) {
-	rootPath, err := ioutil.TempDir(globalTestTmpDir, "minio-")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(rootPath)
+	rootPath := t.TempDir()
 
 	globalConfigDir = &ConfigDir{path: rootPath}
 	configPath := rootPath + SlashSeparator + minioConfigFile
@@ -331,35 +313,31 @@ func TestServerConfigMigrateFaultyConfig(t *testing.T) {
 
 // Test if all migrate code returns error with corrupted config files
 func TestServerConfigMigrateCorruptedConfig(t *testing.T) {
-	rootPath, err := ioutil.TempDir(globalTestTmpDir, "minio-")
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(rootPath)
+	rootPath := t.TempDir()
 
 	globalConfigDir = &ConfigDir{path: rootPath}
 	configPath := rootPath + SlashSeparator + minioConfigFile
 
 	for i := 3; i <= 17; i++ {
 		// Create a corrupted config file
-		if err = ioutil.WriteFile(configPath, []byte(fmt.Sprintf("{ \"version\":\"%d\", \"credential\": { \"accessKey\": 1 } }", i)),
+		if err := ioutil.WriteFile(configPath, []byte(fmt.Sprintf("{ \"version\":\"%d\", \"credential\": { \"accessKey\": 1 } }", i)),
 			0o644); err != nil {
 			t.Fatal("Unexpected error: ", err)
 		}
 
 		// Test different migrate versions and be sure they are returning an error
-		if err = migrateConfig(); err == nil {
+		if err := migrateConfig(); err == nil {
 			t.Fatal("migrateConfig() should fail with a corrupted json")
 		}
 	}
 
 	// Create a corrupted config file for version '2'.
-	if err = ioutil.WriteFile(configPath, []byte("{ \"version\":\"2\", \"credentials\": { \"accessKeyId\": 1 } }"), 0o644); err != nil {
+	if err := ioutil.WriteFile(configPath, []byte("{ \"version\":\"2\", \"credentials\": { \"accessKeyId\": 1 } }"), 0o644); err != nil {
 		t.Fatal("Unexpected error: ", err)
 	}
 
 	// Test different migrate versions and be sure they are returning an error
-	if err = migrateConfig(); err == nil {
+	if err := migrateConfig(); err == nil {
 		t.Fatal("migrateConfig() should fail with a corrupted json")
 	}
 }

cmd/consolelogger.go

@@ -18,10 +18,11 @@
 package cmd
 
 import (
-	ring "container/ring"
+	"container/ring"
 	"context"
 	"sync"
 
+	"github.com/minio/madmin-go"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/minio/internal/logger/message/log"
 	"github.com/minio/minio/internal/logger/target/console"
@@ -71,11 +72,11 @@ func (sys *HTTPConsoleLoggerSys) SetNodeName(nodeName string) {
 // HasLogListeners returns true if console log listeners are registered
 // for this node or peers
 func (sys *HTTPConsoleLoggerSys) HasLogListeners() bool {
-	return sys != nil && sys.pubsub.NumSubscribers() > 0
+	return sys != nil && sys.pubsub.NumSubscribers(madmin.LogMaskAll) > 0
 }
 
 // Subscribe starts console logging for this node.
-func (sys *HTTPConsoleLoggerSys) Subscribe(subCh chan interface{}, doneCh <-chan struct{}, node string, last int, logKind string, filter func(entry interface{}) bool) error {
+func (sys *HTTPConsoleLoggerSys) Subscribe(subCh chan pubsub.Maskable, doneCh <-chan struct{}, node string, last int, logKind madmin.LogMask, filter func(entry pubsub.Maskable) bool) error {
 	// Enable console logging for remote client.
 	if !sys.HasLogListeners() {
 		logger.AddSystemTarget(sys)
@@ -115,8 +116,7 @@ func (sys *HTTPConsoleLoggerSys) Subscribe(subCh chan interface{}, doneCh <-chan
 			}
 		}
 	}
-
-	return sys.pubsub.Subscribe(subCh, doneCh, filter)
+	return sys.pubsub.Subscribe(pubsub.MaskFromMaskable(madmin.LogMaskAll), subCh, doneCh, filter)
 }
 
 // Init if HTTPConsoleLoggerSys is valid, always returns nil right now
@@ -163,9 +163,9 @@ func (sys *HTTPConsoleLoggerSys) Type() types.TargetType {
 
 // Send log message 'e' to console and publish to console
 // log pubsub system
-func (sys *HTTPConsoleLoggerSys) Send(e interface{}, logKind string) error {
+func (sys *HTTPConsoleLoggerSys) Send(entry interface{}) error {
 	var lg log.Info
-	switch e := e.(type) {
+	switch e := entry.(type) {
 	case log.Entry:
 		lg = log.Info{Entry: e, NodeName: sys.nodeName}
 	case string:
@@ -179,5 +179,5 @@ func (sys *HTTPConsoleLoggerSys) Send(e interface{}, logKind string) error {
 	sys.logBuf = sys.logBuf.Next()
 	sys.Unlock()
 
-	return sys.console.Send(e, string(logger.All))
+	return sys.console.Send(entry, string(logger.All))
 }

cmd/data-scanner-metric.go

@@ -0,0 +1,315 @@
+package cmd
+
+import (
+	"strings"
+	"sync"
+	"sync/atomic"
+	"time"
+	"unsafe"
+
+	"github.com/minio/madmin-go"
+	"github.com/minio/minio/internal/bucket/lifecycle"
+)
+
+//go:generate stringer -type=scannerMetric -trimprefix=scannerMetric $GOFILE
+
+type scannerMetric uint8
+
+type scannerMetrics struct {
+	// All fields must be accessed atomically and aligned.
+	operations [scannerMetricLast]uint64
+	latency    [scannerMetricLastRealtime]lockedLastMinuteLatency
+
+	// actions records actions performed.
+	actions        [lifecycle.ActionCount]uint64
+	actionsLatency [lifecycle.ActionCount]lockedLastMinuteLatency
+
+	// currentPaths contains (string,*currentPathTracker) for each disk processing.
+	// Alignment not required.
+	currentPaths sync.Map
+
+	cycleInfoMu sync.Mutex
+	cycleInfo   *currentScannerCycle
+}
+
+var globalScannerMetrics scannerMetrics
+
+const (
+	// START Realtime metrics, that only to records
+	// last minute latencies and total operation count.
+	scannerMetricReadMetadata scannerMetric = iota
+	scannerMetricCheckMissing
+	scannerMetricSaveUsage
+	scannerMetricApplyAll
+	scannerMetricApplyVersion
+	scannerMetricTierObjSweep
+	scannerMetricHealCheck
+	scannerMetricILM
+	scannerMetricCheckReplication
+	scannerMetricYield
+
+	// START Trace metrics:
+	scannerMetricStartTrace
+	scannerMetricScanObject // Scan object. All operations included.
+
+	// END realtime metrics:
+	scannerMetricLastRealtime
+
+	// Trace only metrics:
+	scannerMetricScanFolder     // Scan a folder on disk, recursively.
+	scannerMetricScanCycle      // Full cycle, cluster global
+	scannerMetricScanBucketDisk // Single bucket on one disk
+
+	// Must be last:
+	scannerMetricLast
+)
+
+// log scanner action.
+// Use for s > scannerMetricStartTrace
+func (p *scannerMetrics) log(s scannerMetric, paths ...string) func() {
+	startTime := time.Now()
+	return func() {
+		duration := time.Since(startTime)
+
+		atomic.AddUint64(&p.operations[s], 1)
+		if s < scannerMetricLastRealtime {
+			p.latency[s].add(duration)
+		}
+
+		if s > scannerMetricStartTrace && globalTrace.NumSubscribers(madmin.TraceScanner) > 0 {
+			globalTrace.Publish(scannerTrace(s, startTime, duration, strings.Join(paths, " ")))
+		}
+	}
+}
+
+// time a scanner action.
+// Use for s < scannerMetricLastRealtime
+func (p *scannerMetrics) time(s scannerMetric) func() {
+	startTime := time.Now()
+	return func() {
+		duration := time.Since(startTime)
+
+		atomic.AddUint64(&p.operations[s], 1)
+		if s < scannerMetricLastRealtime {
+			p.latency[s].add(duration)
+		}
+	}
+}
+
+// timeSize add time and size of a scanner action.
+// Use for s < scannerMetricLastRealtime
+func (p *scannerMetrics) timeSize(s scannerMetric) func(sz int) {
+	startTime := time.Now()
+	return func(sz int) {
+		duration := time.Since(startTime)
+
+		atomic.AddUint64(&p.operations[s], 1)
+		if s < scannerMetricLastRealtime {
+			p.latency[s].addSize(duration, int64(sz))
+		}
+	}
+}
+
+// incTime will increment time on metric s with a specific duration.
+// Use for s < scannerMetricLastRealtime
+func (p *scannerMetrics) incTime(s scannerMetric, d time.Duration) {
+	atomic.AddUint64(&p.operations[s], 1)
+	if s < scannerMetricLastRealtime {
+		p.latency[s].add(d)
+	}
+}
+
+func (p *scannerMetrics) incNoTime(s scannerMetric) {
+	atomic.AddUint64(&p.operations[s], 1)
+	if s < scannerMetricLastRealtime {
+		p.latency[s].add(0)
+	}
+}
+
+// timeILM times an ILM action.
+// lifecycle.NoneAction is ignored.
+// Use for s < scannerMetricLastRealtime
+func (p *scannerMetrics) timeILM(a lifecycle.Action) func() {
+	if a == lifecycle.NoneAction || a >= lifecycle.ActionCount {
+		return func() {}
+	}
+	startTime := time.Now()
+	return func() {
+		duration := time.Since(startTime)
+		atomic.AddUint64(&p.actions[a], 1)
+		p.actionsLatency[a].add(duration)
+	}
+}
+
+type currentPathTracker struct {
+	name *unsafe.Pointer // contains atomically accessed *string
+}
+
+// currentPathUpdater provides a lightweight update function for keeping track of
+// current objects for each disk.
+// Returns a function that can be used to update the current object
+// and a function to call to when processing finished.
+func (p *scannerMetrics) currentPathUpdater(disk, initial string) (update func(path string), done func()) {
+	initialPtr := unsafe.Pointer(&initial)
+	tracker := &currentPathTracker{
+		name: &initialPtr,
+	}
+
+	p.currentPaths.Store(disk, tracker)
+	return func(path string) {
+			atomic.StorePointer(tracker.name, unsafe.Pointer(&path))
+		}, func() {
+			p.currentPaths.Delete(disk)
+		}
+}
+
+// getCurrentPaths returns the paths currently being processed.
+func (p *scannerMetrics) getCurrentPaths() []string {
+	var res []string
+	prefix := globalLocalNodeName + "/"
+	p.currentPaths.Range(func(key, value interface{}) bool {
+		// We are a bit paranoid, but better miss an entry than crash.
+		name, ok := key.(string)
+		if !ok {
+			return true
+		}
+		obj, ok := value.(*currentPathTracker)
+		if !ok {
+			return true
+		}
+		strptr := (*string)(atomic.LoadPointer(obj.name))
+		if strptr != nil {
+			res = append(res, pathJoin(prefix, name, *strptr))
+		}
+		return true
+	})
+	return res
+}
+
+// activeDisks returns the number of currently active disks.
+// (since this is concurrent it may not be 100% reliable)
+func (p *scannerMetrics) activeDisks() int {
+	var i int
+	p.currentPaths.Range(func(k, v interface{}) bool {
+		i++
+		return true
+	})
+	return i
+}
+
+// lifetime returns the lifetime count of the specified metric.
+func (p *scannerMetrics) lifetime(m scannerMetric) uint64 {
+	if m < 0 || m >= scannerMetricLast {
+		return 0
+	}
+	val := atomic.LoadUint64(&p.operations[m])
+	return val
+}
+
+// lastMinute returns the last minute statistics of a metric.
+// m should be < scannerMetricLastRealtime
+func (p *scannerMetrics) lastMinute(m scannerMetric) AccElem {
+	if m < 0 || m >= scannerMetricLastRealtime {
+		return AccElem{}
+	}
+	val := p.latency[m].total()
+	return val
+}
+
+// lifetimeActions returns the lifetime count of the specified ilm metric.
+func (p *scannerMetrics) lifetimeActions(a lifecycle.Action) uint64 {
+	if a == lifecycle.NoneAction || a >= lifecycle.ActionCount {
+		return 0
+	}
+	val := atomic.LoadUint64(&p.actions[a])
+	return val
+}
+
+// lastMinuteActions returns the last minute statistics of an ilm metric.
+func (p *scannerMetrics) lastMinuteActions(a lifecycle.Action) AccElem {
+	if a == lifecycle.NoneAction || a >= lifecycle.ActionCount {
+		return AccElem{}
+	}
+	val := p.actionsLatency[a].total()
+	return val
+}
+
+// setCycle updates the current cycle metrics.
+func (p *scannerMetrics) setCycle(c *currentScannerCycle) {
+	if c != nil {
+		c2 := c.clone()
+		c = &c2
+	}
+	p.cycleInfoMu.Lock()
+	p.cycleInfo = c
+	p.cycleInfoMu.Unlock()
+}
+
+// getCycle returns the current cycle metrics.
+// If not nil, the returned value can safely be modified.
+func (p *scannerMetrics) getCycle() *currentScannerCycle {
+	p.cycleInfoMu.Lock()
+	defer p.cycleInfoMu.Unlock()
+	if p.cycleInfo == nil {
+		return nil
+	}
+	c := p.cycleInfo.clone()
+	return &c
+}
+
+func (p *scannerMetrics) report() madmin.ScannerMetrics {
+	var m madmin.ScannerMetrics
+	cycle := p.getCycle()
+	if cycle != nil {
+		m.CurrentCycle = cycle.current
+		m.CyclesCompletedAt = cycle.cycleCompleted
+		m.CurrentStarted = cycle.started
+	}
+	m.CollectedAt = time.Now()
+	m.ActivePaths = p.getCurrentPaths()
+	m.LifeTimeOps = make(map[string]uint64, scannerMetricLast)
+	for i := scannerMetric(0); i < scannerMetricLast; i++ {
+		if n := atomic.LoadUint64(&p.operations[i]); n > 0 {
+			m.LifeTimeOps[i.String()] = n
+		}
+	}
+	if len(m.LifeTimeOps) == 0 {
+		m.LifeTimeOps = nil
+	}
+
+	m.LastMinute.Actions = make(map[string]madmin.TimedAction, scannerMetricLastRealtime)
+	for i := scannerMetric(0); i < scannerMetricLastRealtime; i++ {
+		lm := p.lastMinute(i)
+		if lm.N > 0 {
+			m.LastMinute.Actions[i.String()] = lm.asTimedAction()
+		}
+	}
+	if len(m.LastMinute.Actions) == 0 {
+		m.LastMinute.Actions = nil
+	}
+
+	// ILM
+	m.LifeTimeILM = make(map[string]uint64)
+	for i := lifecycle.NoneAction + 1; i < lifecycle.ActionCount; i++ {
+		if n := atomic.LoadUint64(&p.actions[i]); n > 0 {
+			m.LifeTimeILM[i.String()] = n
+		}
+	}
+	if len(m.LifeTimeILM) == 0 {
+		m.LifeTimeILM = nil
+	}
+
+	if len(m.LifeTimeILM) > 0 {
+		m.LastMinute.ILM = make(map[string]madmin.TimedAction, len(m.LifeTimeILM))
+		for i := lifecycle.NoneAction + 1; i < lifecycle.ActionCount; i++ {
+			lm := p.lastMinuteActions(i)
+			if lm.N > 0 {
+				m.LastMinute.ILM[i.String()] = madmin.TimedAction{Count: uint64(lm.N), AccTime: uint64(lm.Total)}
+			}
+		}
+		if len(m.LastMinute.ILM) == 0 {
+			m.LastMinute.ILM = nil
+		}
+	}
+	return m
+}

cmd/data-scanner.go

@@ -31,7 +31,6 @@ import (
 	"path"
 	"strings"
 	"sync"
-	"sync/atomic"
 	"time"
 
 	"github.com/bits-and-blooms/bloom/v3"
@@ -49,12 +48,13 @@ import (
 )
 
 const (
-	dataScannerSleepPerFolder     = time.Millisecond                 // Time to wait between folders.
-	dataUsageUpdateDirCycles      = 16                               // Visit all folders every n cycles.
-	dataScannerCompactLeastObject = 500                              // Compact when there is less than this many objects in a branch.
-	dataScannerCompactAtChildren  = 10000                            // Compact when there are this many children in a branch.
-	dataScannerCompactAtFolders   = dataScannerCompactAtChildren / 4 // Compact when this many subfolders in a single folder.
-	dataScannerStartDelay         = 1 * time.Minute                  // Time to wait on startup and between cycles.
+	dataScannerSleepPerFolder        = time.Millisecond                 // Time to wait between folders.
+	dataUsageUpdateDirCycles         = 16                               // Visit all folders every n cycles.
+	dataScannerCompactLeastObject    = 500                              // Compact when there is less than this many objects in a branch.
+	dataScannerCompactAtChildren     = 10000                            // Compact when there are this many children in a branch.
+	dataScannerCompactAtFolders      = dataScannerCompactAtChildren / 4 // Compact when this many subfolders in a single folder.
+	dataScannerForceCompactAtFolders = 1_000_000                        // Compact when this many subfolders in a single folder (even top level).
+	dataScannerStartDelay            = 1 * time.Minute                  // Time to wait on startup and between cycles.
 
 	healDeleteDangling    = true
 	healFolderIncludeProb = 32  // Include a clean folder one in n cycles.
@@ -64,9 +64,13 @@ const (
 var (
 	globalHealConfig heal.Config
 
-	dataScannerLeaderLockTimeout = newDynamicTimeout(30*time.Second, 10*time.Second)
+	dataScannerLeaderLockTimeout = newDynamicTimeoutWithOpts(dynamicTimeoutOpts{
+		timeout:       30 * time.Second,
+		minimum:       10 * time.Second,
+		retryInterval: time.Second,
+	})
 	// Sleeper values are updated when config is loaded.
-	scannerSleeper = newDynamicSleeper(10, 10*time.Second)
+	scannerSleeper = newDynamicSleeper(10, 10*time.Second, true)
 	scannerCycle   = uatomic.NewDuration(dataScannerStartDelay)
 )
 
@@ -162,35 +166,45 @@ func runDataScanner(pctx context.Context, objAPI ObjectLayer) {
 	// No unlock for "leader" lock.
 
 	// Load current bloom cycle
-	nextBloomCycle := intDataUpdateTracker.current() + 1
+	var cycleInfo currentScannerCycle
+	cycleInfo.next = intDataUpdateTracker.current() + 1
 
 	buf, _ := readConfig(ctx, objAPI, dataUsageBloomNamePath)
-	if len(buf) >= 8 {
-		if err = binary.Read(bytes.NewReader(buf), binary.LittleEndian, &nextBloomCycle); err != nil {
-			logger.LogIf(ctx, err)
-		}
+	if len(buf) == 8 {
+		cycleInfo.next = binary.LittleEndian.Uint64(buf)
+	} else if len(buf) > 8 {
+		cycleInfo.next = binary.LittleEndian.Uint64(buf[:8])
+		buf = buf[8:]
+		_, err = cycleInfo.UnmarshalMsg(buf)
+		logger.LogIf(ctx, err)
 	}
 
 	scannerTimer := time.NewTimer(scannerCycle.Load())
 	defer scannerTimer.Stop()
+	defer globalScannerMetrics.setCycle(nil)
 
 	for {
 		select {
 		case <-ctx.Done():
 			return
 		case <-scannerTimer.C:
-			if intDataUpdateTracker.debug {
-				console.Debugln("starting scanner cycle")
-			}
+			// Reset the timer for next cycle.
+			// If scanner takes longer we start at once.
+			scannerTimer.Reset(scannerCycle.Load())
+
+			stopFn := globalScannerMetrics.log(scannerMetricScanCycle)
+			cycleInfo.current = cycleInfo.next
+			cycleInfo.started = time.Now()
+			globalScannerMetrics.setCycle(&cycleInfo)
 
 			bgHealInfo := readBackgroundHealInfo(ctx, objAPI)
-			scanMode := getCycleScanMode(nextBloomCycle, bgHealInfo.BitrotStartCycle, bgHealInfo.BitrotStartTime)
+			scanMode := getCycleScanMode(cycleInfo.current, bgHealInfo.BitrotStartCycle, bgHealInfo.BitrotStartTime)
 			if bgHealInfo.CurrentScanMode != scanMode {
 				newHealInfo := bgHealInfo
 				newHealInfo.CurrentScanMode = scanMode
 				if scanMode == madmin.HealDeepScan {
 					newHealInfo.BitrotStartTime = time.Now().UTC()
-					newHealInfo.BitrotStartCycle = nextBloomCycle
+					newHealInfo.BitrotStartCycle = cycleInfo.current
 				}
 				saveBackgroundHealInfo(ctx, objAPI, newHealInfo)
 			}
@@ -198,22 +212,27 @@ func runDataScanner(pctx context.Context, objAPI ObjectLayer) {
 			// Wait before starting next cycle and wait on startup.
 			results := make(chan DataUsageInfo, 1)
 			go storeDataUsageInBackend(ctx, objAPI, results)
-			bf, err := globalNotificationSys.updateBloomFilter(ctx, nextBloomCycle)
+			bf, err := globalNotificationSys.updateBloomFilter(ctx, cycleInfo.current)
 			logger.LogIf(ctx, err)
-			err = objAPI.NSScanner(ctx, bf, results, uint32(nextBloomCycle), scanMode)
+			err = objAPI.NSScanner(ctx, bf, results, uint32(cycleInfo.current), scanMode)
 			logger.LogIf(ctx, err)
+			stopFn()
 			if err == nil {
 				// Store new cycle...
-				nextBloomCycle++
-				var tmp [8]byte
-				binary.LittleEndian.PutUint64(tmp[:], nextBloomCycle)
-				if err = saveConfig(ctx, objAPI, dataUsageBloomNamePath, tmp[:]); err != nil {
-					logger.LogIf(ctx, err)
+				cycleInfo.next++
+				cycleInfo.current = 0
+				cycleInfo.cycleCompleted = append(cycleInfo.cycleCompleted, time.Now())
+				if len(cycleInfo.cycleCompleted) > dataUsageUpdateDirCycles {
+					cycleInfo.cycleCompleted = cycleInfo.cycleCompleted[len(cycleInfo.cycleCompleted)-dataUsageUpdateDirCycles:]
 				}
+				globalScannerMetrics.setCycle(&cycleInfo)
+				tmp := make([]byte, 8, 8+cycleInfo.Msgsize())
+				// Cycle for backward compat.
+				binary.LittleEndian.PutUint64(tmp, cycleInfo.next)
+				tmp, _ = cycleInfo.MarshalMsg(tmp)
+				err = saveConfig(ctx, objAPI, dataUsageBloomNamePath, tmp)
+				logger.LogIf(ctx, err)
 			}
-
-			// Reset the timer for next cycle.
-			scannerTimer.Reset(scannerCycle.Load())
 		}
 	}
 }
@@ -244,24 +263,11 @@ type folderScanner struct {
 	// Will not be closed when returned.
 	updates    chan<- dataUsageEntry
 	lastUpdate time.Time
+
+	// updateCurrentPath should be called whenever a new path is scanned.
+	updateCurrentPath func(string)
 }
 
-type scannerStats struct {
-	// All fields must be accessed atomically and aligned.
-
-	accTotalObjects  uint64
-	accTotalVersions uint64
-	accFolders       uint64
-	bucketsStarted   uint64
-	bucketsFinished  uint64
-	ilmChecks        uint64
-
-	// actions records actions performed.
-	actions [lifecycle.ActionCount]uint64
-}
-
-var globalScannerStats scannerStats
-
 // Cache structure and compaction:
 //
 // A cache structure will be kept with a tree of usages.
@@ -305,24 +311,14 @@ var globalScannerStats scannerStats
 // Before each operation sleepDuration is called which can be used to temporarily halt the scanner.
 // If the supplied context is canceled the function will return at the first chance.
 func scanDataFolder(ctx context.Context, poolIdx, setIdx int, basePath string, cache dataUsageCache, getSize getSizeFn, scanMode madmin.HealScanMode) (dataUsageCache, error) {
-	t := UTCNow()
-
 	logPrefix := color.Green("data-usage: ")
-	logSuffix := color.Blue("- %v + %v", basePath, cache.Info.Name)
-	atomic.AddUint64(&globalScannerStats.bucketsStarted, 1)
-	defer func() {
-		atomic.AddUint64(&globalScannerStats.bucketsFinished, 1)
-	}()
-	if intDataUpdateTracker.debug {
-		defer func() {
-			console.Debugf(logPrefix+" Scanner time: %v %s\n", time.Since(t), logSuffix)
-		}()
-	}
 
 	switch cache.Info.Name {
 	case "", dataUsageRoot:
 		return cache, errors.New("internal error: root scan attempted")
 	}
+	updatePath, closeDisk := globalScannerMetrics.currentPathUpdater(basePath, cache.Info.Name)
+	defer closeDisk()
 
 	s := folderScanner{
 		root:                  basePath,
@@ -335,6 +331,7 @@ func scanDataFolder(ctx context.Context, poolIdx, setIdx int, basePath string, c
 		healObjectSelect:      0,
 		scanMode:              scanMode,
 		updates:               cache.Info.updates,
+		updateCurrentPath:     updatePath,
 	}
 
 	// Add disks for set healing.
@@ -366,14 +363,8 @@ func scanDataFolder(ctx context.Context, poolIdx, setIdx int, basePath string, c
 			s.withFilter = nil
 		}
 	}
-	if s.dataUsageScannerDebug {
-		console.Debugf(logPrefix+"Start scanning. Bloom filter: %v %s\n", s.withFilter != nil, logSuffix)
-	}
 
 	done := ctx.Done()
-	if s.dataUsageScannerDebug {
-		console.Debugf(logPrefix+"Cycle: %v, Entries: %v %s\n", cache.Info.NextCycle, len(cache.Cache), logSuffix)
-	}
 
 	// Read top level in bucket.
 	select {
@@ -389,9 +380,6 @@ func scanDataFolder(ctx context.Context, poolIdx, setIdx int, basePath string, c
 		return cache, err
 	}
 
-	if s.dataUsageScannerDebug {
-		console.Debugf(logPrefix+"Finished scanner, %v entries (%+v) %s \n", len(s.newCache.Cache), *s.newCache.sizeRecursive(s.newCache.Info.Name), logSuffix)
-	}
 	s.newCache.Info.LastUpdate = UTCNow()
 	s.newCache.Info.NextCycle = cache.Info.NextCycle
 	return s.newCache, nil
@@ -420,10 +408,10 @@ func (f *folderScanner) sendUpdate() {
 func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, into *dataUsageEntry) error {
 	done := ctx.Done()
 	scannerLogPrefix := color.Green("folder-scanner:")
+
 	thisHash := hashPath(folder.name)
 	// Store initial compaction state.
 	wasCompacted := into.Compacted
-	atomic.AddUint64(&globalScannerStats.accFolders, 1)
 
 	for {
 		select {
@@ -583,9 +571,11 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 		}
 
 		// If we have many subfolders, compact ourself.
-		if !into.Compacted &&
-			f.newCache.Info.Name != folder.name &&
-			len(existingFolders)+len(newFolders) >= dataScannerCompactAtFolders {
+		shouldCompact := f.newCache.Info.Name != folder.name &&
+			len(existingFolders)+len(newFolders) >= dataScannerCompactAtFolders ||
+			len(existingFolders)+len(newFolders) >= dataScannerForceCompactAtFolders
+
+		if !into.Compacted && shouldCompact {
 			into.Compacted = true
 			newFolders = append(newFolders, existingFolders...)
 			existingFolders = nil
@@ -648,7 +638,10 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 					f.updateCache.replaceHashed(h, &thisHash, dataUsageEntry{})
 				}
 			}
+			f.updateCurrentPath(folder.name)
+			stopFn := globalScannerMetrics.log(scannerMetricScanFolder, f.root, folder.name)
 			scanFolder(folder)
+			stopFn()
 			// Add new folders if this is new and we don't have existing.
 			if !into.Compacted {
 				parent := f.updateCache.find(thisHash.Key())
@@ -676,7 +669,10 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 					folder.objectHealProbDiv = f.healFolderInclude
 				}
 			}
+			f.updateCurrentPath(folder.name)
+			stopFn := globalScannerMetrics.log(scannerMetricScanFolder, f.root, folder.name, "EXISTING")
 			scanFolder(folder)
+			stopFn()
 		}
 
 		// Scan for healing
@@ -717,9 +713,8 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 		healObjectsPrefix := color.Green("healObjects:")
 		for k := range abandonedChildren {
 			bucket, prefix := path2BucketObject(k)
-			if f.dataUsageScannerDebug {
-				console.Debugf(scannerLogPrefix+" checking disappeared folder: %v/%v\n", bucket, prefix)
-			}
+			stopFn := globalScannerMetrics.time(scannerMetricCheckMissing)
+			f.updateCurrentPath(k)
 
 			if bucket != resolver.bucket {
 				// Bucket might be missing as well with abandoned children.
@@ -746,11 +741,7 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 					}
 				},
 				// Some disks have data for this.
-				partial: func(entries metaCacheEntries, nAgreed int, errs []error) {
-					if f.dataUsageScannerDebug {
-						console.Debugf(healObjectsPrefix+" got partial, %d agreed, errs: %v\n", nAgreed, errs)
-					}
-
+				partial: func(entries metaCacheEntries, errs []error) {
 					entry, ok := entries.resolve(&resolver)
 					if !ok {
 						// check if we can get one entry atleast
@@ -807,6 +798,7 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 				},
 			})
 
+			stopFn()
 			if f.dataUsageScannerDebug && err != nil && err != errFileNotFound {
 				console.Debugf(healObjectsPrefix+" checking returned value %v (%T)\n", err, err)
 			}
@@ -814,7 +806,9 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 			// Add unless healing returned an error.
 			if foundObjs {
 				this := cachedFolder{name: k, parent: &thisHash, objectHealProbDiv: 1}
+				stopFn := globalScannerMetrics.log(scannerMetricScanFolder, f.root, this.name, "HEALED")
 				scanFolder(this)
+				stopFn()
 			}
 		}
 		break
@@ -964,7 +958,6 @@ func (i *scannerItem) applyLifecycle(ctx context.Context, o ObjectLayer, oi Obje
 		return false, size
 	}
 
-	atomic.AddUint64(&globalScannerStats.ilmChecks, 1)
 	versionID := oi.VersionID
 	rCfg, _ := globalBucketObjectLockSys.Get(i.bucket)
 	action := evalActionFromLifecycle(ctx, *i.lifeCycle, rCfg, oi, false)
@@ -975,7 +968,7 @@ func (i *scannerItem) applyLifecycle(ctx context.Context, o ObjectLayer, oi Obje
 			console.Debugf(applyActionsLogPrefix+" lifecycle: %q Initial scan: %v\n", i.objectPath(), action)
 		}
 	}
-	atomic.AddUint64(&globalScannerStats.actions[action], 1)
+	defer globalScannerMetrics.timeILM(action)
 
 	switch action {
 	case lifecycle.DeleteAction, lifecycle.DeleteVersionAction, lifecycle.DeleteRestoredAction, lifecycle.DeleteRestoredVersionAction:
@@ -1092,16 +1085,23 @@ func (i *scannerItem) applyVersionActions(ctx context.Context, o ObjectLayer, fi
 // The metadata will be compared to consensus on the object layer before any changes are applied.
 // If no metadata is supplied, -1 is returned if no action is taken.
 func (i *scannerItem) applyActions(ctx context.Context, o ObjectLayer, oi ObjectInfo, sizeS *sizeSummary) int64 {
+	done := globalScannerMetrics.time(scannerMetricILM)
 	applied, size := i.applyLifecycle(ctx, o, oi)
+	done()
+
 	// For instance, an applied lifecycle means we remove/transitioned an object
 	// from the current deployment, which means we don't have to call healing
 	// routine even if we are asked to do via heal flag.
 	if !applied {
 		if i.heal.enabled {
+			done := globalScannerMetrics.time(scannerMetricHealCheck)
 			size = i.applyHealing(ctx, o, oi)
+			done()
 		}
 		// replicate only if lifecycle rules are not applied.
+		done := globalScannerMetrics.time(scannerMetricCheckReplication)
 		i.healReplication(ctx, o, oi.Clone(), sizeS)
+		done()
 	}
 	return size
 }
@@ -1229,29 +1229,16 @@ func (i *scannerItem) objectPath() string {
 
 // healReplication will heal a scanned item that has failed replication.
 func (i *scannerItem) healReplication(ctx context.Context, o ObjectLayer, oi ObjectInfo, sizeS *sizeSummary) {
-	roi := getHealReplicateObjectInfo(oi, i.replication)
-	if !roi.Dsc.ReplicateAny() {
+	if oi.VersionID == "" {
 		return
 	}
-
+	if i.replication.Config == nil {
+		return
+	}
+	roi := queueReplicationHeal(ctx, oi.Bucket, oi, i.replication)
 	if oi.DeleteMarker || !oi.VersionPurgeStatus.Empty() {
-		// heal delete marker replication failure or versioned delete replication failure
-		if oi.ReplicationStatus == replication.Pending ||
-			oi.ReplicationStatus == replication.Failed ||
-			oi.VersionPurgeStatus == Failed || oi.VersionPurgeStatus == Pending {
-			i.healReplicationDeletes(ctx, o, roi)
-			return
-		}
-		// if replication status is Complete on DeleteMarker and existing object resync required
-		if roi.ExistingObjResync.mustResync() && (oi.ReplicationStatus == replication.Completed || oi.ReplicationStatus.Empty()) {
-			i.healReplicationDeletes(ctx, o, roi)
-			return
-		}
 		return
 	}
-	if roi.ExistingObjResync.mustResync() {
-		roi.OpType = replication.ExistingObjectReplicationType
-	}
 
 	if sizeS.replTargetStats == nil && len(roi.TargetStatuses) > 0 {
 		sizeS.replTargetStats = make(map[string]replTargetSizeSummary)
@@ -1281,48 +1268,9 @@ func (i *scannerItem) healReplication(ctx context.Context, o ObjectLayer, oi Obj
 	}
 
 	switch oi.ReplicationStatus {
-	case replication.Pending, replication.Failed:
-		globalReplicationPool.queueReplicaTask(roi)
-		return
 	case replication.Replica:
 		sizeS.replicaSize += oi.Size
 	}
-	if roi.ExistingObjResync.mustResync() {
-		globalReplicationPool.queueReplicaTask(roi)
-	}
-}
-
-// healReplicationDeletes will heal a scanned deleted item that failed to replicate deletes.
-func (i *scannerItem) healReplicationDeletes(ctx context.Context, o ObjectLayer, roi ReplicateObjectInfo) {
-	// handle soft delete and permanent delete failures here.
-	if roi.DeleteMarker || !roi.VersionPurgeStatus.Empty() {
-		versionID := ""
-		dmVersionID := ""
-		if roi.VersionPurgeStatus.Empty() {
-			dmVersionID = roi.VersionID
-		} else {
-			versionID = roi.VersionID
-		}
-
-		doi := DeletedObjectReplicationInfo{
-			DeletedObject: DeletedObject{
-				ObjectName:            roi.Name,
-				DeleteMarkerVersionID: dmVersionID,
-				VersionID:             versionID,
-				ReplicationState:      roi.getReplicationState(roi.Dsc.String(), versionID, true),
-				DeleteMarkerMTime:     DeleteMarkerMTime{roi.ModTime},
-				DeleteMarker:          roi.DeleteMarker,
-			},
-			Bucket: roi.Bucket,
-			OpType: replication.HealReplicationType,
-		}
-		if roi.ExistingObjResync.mustResync() {
-			doi.OpType = replication.ExistingObjectReplicationType
-			queueReplicateDeletesWrapper(doi, roi.ExistingObjResync)
-			return
-		}
-		globalReplicationPool.queueReplicaDeleteTask(doi)
-	}
 }
 
 type dynamicSleeper struct {
@@ -1341,15 +1289,20 @@ type dynamicSleeper struct {
 
 	// cycle will be closed
 	cycle chan struct{}
+
+	// isScanner should be set when this is used by the scanner
+	// to record metrics.
+	isScanner bool
 }
 
 // newDynamicSleeper
-func newDynamicSleeper(factor float64, maxWait time.Duration) *dynamicSleeper {
+func newDynamicSleeper(factor float64, maxWait time.Duration, isScanner bool) *dynamicSleeper {
 	return &dynamicSleeper{
-		factor:   factor,
-		cycle:    make(chan struct{}),
-		maxSleep: maxWait,
-		minSleep: 100 * time.Microsecond,
+		factor:    factor,
+		cycle:     make(chan struct{}),
+		maxSleep:  maxWait,
+		minSleep:  100 * time.Microsecond,
+		isScanner: isScanner,
 	}
 }
 
@@ -1379,15 +1332,24 @@ func (d *dynamicSleeper) Timer(ctx context.Context) func() {
 			select {
 			case <-ctx.Done():
 				if !timer.Stop() {
+					if d.isScanner {
+						globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+					}
 					<-timer.C
 				}
 				return
 			case <-timer.C:
+				if d.isScanner {
+					globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+				}
 				return
 			case <-cycle:
 				if !timer.Stop() {
 					// We expired.
 					<-timer.C
+					if d.isScanner {
+						globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+					}
 					return
 				}
 			}
@@ -1418,14 +1380,23 @@ func (d *dynamicSleeper) Sleep(ctx context.Context, base time.Duration) {
 		case <-ctx.Done():
 			if !timer.Stop() {
 				<-timer.C
+				if d.isScanner {
+					globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+				}
 			}
 			return
 		case <-timer.C:
+			if d.isScanner {
+				globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+			}
 			return
 		case <-cycle:
 			if !timer.Stop() {
 				// We expired.
 				<-timer.C
+				if d.isScanner {
+					globalScannerMetrics.incTime(scannerMetricYield, wantSleep)
+				}
 				return
 			}
 		}
@@ -1457,9 +1428,9 @@ const (
 	ILMTransition = " ilm:transition"
 )
 
-func auditLogLifecycle(ctx context.Context, oi ObjectInfo, trigger string) {
+func auditLogLifecycle(ctx context.Context, oi ObjectInfo, event string) {
 	var apiName string
-	switch trigger {
+	switch event {
 	case ILMExpiry:
 		apiName = "ILMExpiry"
 	case ILMFreeVersionDelete:
@@ -1468,7 +1439,7 @@ func auditLogLifecycle(ctx context.Context, oi ObjectInfo, trigger string) {
 		apiName = "ILMTransition"
 	}
 	auditLogInternal(ctx, oi.Bucket, oi.Name, AuditLogOptions{
-		Trigger:   trigger,
+		Event:     event,
 		APIName:   apiName,
 		VersionID: oi.VersionID,
 	})

cmd/data-update-tracker.go

@@ -227,7 +227,7 @@ func (d *dataUpdateTracker) load(ctx context.Context, drives ...string) {
 	for _, drive := range drives {
 
 		cacheFormatPath := pathJoin(drive, dataUpdateTrackerFilename)
-		f, err := os.Open(cacheFormatPath)
+		f, err := OpenFile(cacheFormatPath, readMode, 0o666)
 		if err != nil {
 			if osIsNotExist(err) {
 				continue

cmd/data-update-tracker_test.go

@@ -21,7 +21,6 @@ import (
 	"bytes"
 	"context"
 	"fmt"
-	"io/ioutil"
 	"math/rand"
 	"os"
 	"path"
@@ -63,7 +62,7 @@ func (t *testingLogger) Type() types.TargetType {
 	return types.TargetHTTP
 }
 
-func (t *testingLogger) Send(entry interface{}, errKind string) error {
+func (t *testingLogger) Send(entry interface{}) error {
 	t.mu.Lock()
 	defer t.mu.Unlock()
 	if t.t == nil {
@@ -75,7 +74,7 @@ func (t *testingLogger) Send(entry interface{}, errKind string) error {
 	}
 
 	t.t.Helper()
-	t.t.Log(e.Level, ":", errKind, e.Message)
+	t.t.Log(e.Level, ":", e.Message)
 	return nil
 }
 
@@ -100,15 +99,11 @@ func TestDataUpdateTracker(t *testing.T) {
 
 	dut.Current.bf = dut.newBloomFilter()
 
-	tmpDir, err := ioutil.TempDir("", "TestDataUpdateTracker")
+	tmpDir := t.TempDir()
+	err := os.MkdirAll(filepath.Dir(filepath.Join(tmpDir, dataUpdateTrackerFilename)), os.ModePerm)
 	if err != nil {
 		t.Fatal(err)
 	}
-	err = os.MkdirAll(filepath.Dir(filepath.Join(tmpDir, dataUpdateTrackerFilename)), os.ModePerm)
-	if err != nil {
-		t.Fatal(err)
-	}
-	defer os.RemoveAll(tmpDir)
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 	dut.start(ctx, tmpDir)

cmd/data-usage-cache.go

@@ -1323,3 +1323,19 @@ func (z dataUsageHashMap) Msgsize() (s int) {
 	}
 	return
 }
+
+//msgp:encode ignore currentScannerCycle
+//msgp:decode ignore currentScannerCycle
+
+type currentScannerCycle struct {
+	current        uint64
+	next           uint64
+	started        time.Time
+	cycleCompleted []time.Time
+}
+
+// clone returns a clone.
+func (z currentScannerCycle) clone() currentScannerCycle {
+	z.cycleCompleted = append(make([]time.Time, 0, len(z.cycleCompleted)), z.cycleCompleted...)
+	return z
+}

cmd/data-usage-cache_gen.go

@@ -3,6 +3,8 @@ package cmd
 // Code generated by github.com/tinylib/msgp DO NOT EDIT.
 
 import (
+	"time"
+
 	"github.com/tinylib/msgp/msgp"
 )
 
@@ -284,6 +286,101 @@ func (z *allTierStats) Msgsize() (s int) {
 	return
 }
 
+// MarshalMsg implements msgp.Marshaler
+func (z *currentScannerCycle) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 4
+	// string "current"
+	o = append(o, 0x84, 0xa7, 0x63, 0x75, 0x72, 0x72, 0x65, 0x6e, 0x74)
+	o = msgp.AppendUint64(o, z.current)
+	// string "next"
+	o = append(o, 0xa4, 0x6e, 0x65, 0x78, 0x74)
+	o = msgp.AppendUint64(o, z.next)
+	// string "started"
+	o = append(o, 0xa7, 0x73, 0x74, 0x61, 0x72, 0x74, 0x65, 0x64)
+	o = msgp.AppendTime(o, z.started)
+	// string "cycleCompleted"
+	o = append(o, 0xae, 0x63, 0x79, 0x63, 0x6c, 0x65, 0x43, 0x6f, 0x6d, 0x70, 0x6c, 0x65, 0x74, 0x65, 0x64)
+	o = msgp.AppendArrayHeader(o, uint32(len(z.cycleCompleted)))
+	for za0001 := range z.cycleCompleted {
+		o = msgp.AppendTime(o, z.cycleCompleted[za0001])
+	}
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *currentScannerCycle) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "current":
+			z.current, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "current")
+				return
+			}
+		case "next":
+			z.next, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "next")
+				return
+			}
+		case "started":
+			z.started, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "started")
+				return
+			}
+		case "cycleCompleted":
+			var zb0002 uint32
+			zb0002, bts, err = msgp.ReadArrayHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "cycleCompleted")
+				return
+			}
+			if cap(z.cycleCompleted) >= int(zb0002) {
+				z.cycleCompleted = (z.cycleCompleted)[:zb0002]
+			} else {
+				z.cycleCompleted = make([]time.Time, zb0002)
+			}
+			for za0001 := range z.cycleCompleted {
+				z.cycleCompleted[za0001], bts, err = msgp.ReadTimeBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "cycleCompleted", za0001)
+					return
+				}
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *currentScannerCycle) Msgsize() (s int) {
+	s = 1 + 8 + msgp.Uint64Size + 5 + msgp.Uint64Size + 8 + msgp.TimeSize + 15 + msgp.ArrayHeaderSize + (len(z.cycleCompleted) * (msgp.TimeSize))
+	return
+}
+
 // DecodeMsg implements msgp.Decodable
 func (z *dataUsageCache) DecodeMsg(dc *msgp.Reader) (err error) {
 	var field []byte

cmd/data-usage-cache_gen_test.go

@@ -122,6 +122,64 @@ func BenchmarkDecodeallTierStats(b *testing.B) {
 	}
 }
 
+func TestMarshalUnmarshalcurrentScannerCycle(t *testing.T) {
+	v := currentScannerCycle{}
+	bts, err := v.MarshalMsg(nil)
+	if err != nil {
+		t.Fatal(err)
+	}
+	left, err := v.UnmarshalMsg(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after UnmarshalMsg(): %q", len(left), left)
+	}
+
+	left, err = msgp.Skip(bts)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(left) > 0 {
+		t.Errorf("%d bytes left over after Skip(): %q", len(left), left)
+	}
+}
+
+func BenchmarkMarshalMsgcurrentScannerCycle(b *testing.B) {
+	v := currentScannerCycle{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgcurrentScannerCycle(b *testing.B) {
+	v := currentScannerCycle{}
+	bts := make([]byte, 0, v.Msgsize())
+	bts, _ = v.MarshalMsg(bts[0:0])
+	b.SetBytes(int64(len(bts)))
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		bts, _ = v.MarshalMsg(bts[0:0])
+	}
+}
+
+func BenchmarkUnmarshalcurrentScannerCycle(b *testing.B) {
+	v := currentScannerCycle{}
+	bts, _ := v.MarshalMsg(nil)
+	b.ReportAllocs()
+	b.SetBytes(int64(len(bts)))
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		_, err := v.UnmarshalMsg(bts)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
 func TestMarshalUnmarshaldataUsageCache(t *testing.T) {
 	v := dataUsageCache{}
 	bts, err := v.MarshalMsg(nil)

cmd/data-usage_test.go

@@ -35,12 +35,8 @@ type usageTestFile struct {
 }
 
 func TestDataUsageUpdate(t *testing.T) {
-	base, err := ioutil.TempDir("", "TestDataUsageUpdate")
-	if err != nil {
-		t.Skip(err)
-	}
+	base := t.TempDir()
 	const bucket = "bucket"
-	defer os.RemoveAll(base)
 	files := []usageTestFile{
 		{name: "rootfile", size: 10000},
 		{name: "rootfile2", size: 10000},
@@ -251,12 +247,8 @@ func TestDataUsageUpdate(t *testing.T) {
 }
 
 func TestDataUsageUpdatePrefix(t *testing.T) {
-	base, err := ioutil.TempDir("", "TestDataUpdateUsagePrefix")
-	if err != nil {
-		t.Skip(err)
-	}
+	base := t.TempDir()
 	scannerSleeper.Update(0, 0)
-	defer os.RemoveAll(base)
 	files := []usageTestFile{
 		{name: "bucket/rootfile", size: 10000},
 		{name: "bucket/rootfile2", size: 10000},
@@ -537,12 +529,8 @@ func generateUsageTestFiles(t *testing.T, base, bucket string, nFolders, nFiles,
 }
 
 func TestDataUsageCacheSerialize(t *testing.T) {
-	base, err := ioutil.TempDir("", "TestDataUsageCacheSerialize")
-	if err != nil {
-		t.Skip(err)
-	}
+	base := t.TempDir()
 	const bucket = "abucket"
-	defer os.RemoveAll(base)
 	files := []usageTestFile{
 		{name: "rootfile", size: 10000},
 		{name: "rootfile2", size: 10000},

cmd/decommetric_string.go

@@ -0,0 +1,25 @@
+// Code generated by "stringer -type=decomMetric -trimprefix=decomMetric erasure-server-pool-decom.go"; DO NOT EDIT.
+
+package cmd
+
+import "strconv"
+
+func _() {
+	// An "invalid array index" compiler error signifies that the constant values have changed.
+	// Re-run the stringer command to generate them again.
+	var x [1]struct{}
+	_ = x[decomMetricDecommissionBucket-0]
+	_ = x[decomMetricDecommissionObject-1]
+	_ = x[decomMetricDecommissionRemoveObject-2]
+}
+
+const _decomMetric_name = "DecommissionBucketDecommissionObjectDecommissionRemoveObject"
+
+var _decomMetric_index = [...]uint8{0, 18, 36, 60}
+
+func (i decomMetric) String() string {
+	if i >= decomMetric(len(_decomMetric_index)-1) {
+		return "decomMetric(" + strconv.FormatInt(int64(i), 10) + ")"
+	}
+	return _decomMetric_name[_decomMetric_index[i]:_decomMetric_index[i+1]]
+}

cmd/disk-cache-backend.go

@@ -624,7 +624,7 @@ func (c *diskCache) saveMetadata(ctx context.Context, bucket, object string, met
 	if err := os.MkdirAll(cachedPath, 0o777); err != nil {
 		return err
 	}
-	f, err := os.OpenFile(metaPath, os.O_RDWR|os.O_CREATE, 0o666)
+	f, err := OpenFile(metaPath, os.O_RDWR|os.O_CREATE|writeMode, 0o666)
 	if err != nil {
 		return err
 	}
@@ -687,7 +687,7 @@ func (c *diskCache) updateMetadata(ctx context.Context, bucket, object, etag str
 	if err := os.MkdirAll(cachedPath, 0o777); err != nil {
 		return err
 	}
-	f, err := os.OpenFile(metaPath, os.O_RDWR, 0o666)
+	f, err := OpenFile(metaPath, os.O_RDWR|writeMode, 0o666)
 	if err != nil {
 		return err
 	}
@@ -797,25 +797,25 @@ func (c *diskCache) bitrotWriteToCache(cachePath, fileName string, reader io.Rea
 	return bytesWritten, base64.StdEncoding.EncodeToString(md5sumCurr), nil
 }
 
-func newCacheEncryptReader(content io.Reader, bucket, object string, metadata map[string]string) (r io.Reader, err error) {
-	objectEncryptionKey, err := newCacheEncryptMetadata(bucket, object, metadata)
+func newCacheEncryptReader(ctx context.Context, content io.Reader, bucket, object string, metadata map[string]string) (r io.Reader, err error) {
+	objectEncryptionKey, err := newCacheEncryptMetadata(ctx, bucket, object, metadata)
 	if err != nil {
 		return nil, err
 	}
 
-	reader, err := sio.EncryptReader(content, sio.Config{Key: objectEncryptionKey, MinVersion: sio.Version20, CipherSuites: fips.CipherSuitesDARE()})
+	reader, err := sio.EncryptReader(content, sio.Config{Key: objectEncryptionKey, MinVersion: sio.Version20, CipherSuites: fips.DARECiphers()})
 	if err != nil {
 		return nil, crypto.ErrInvalidCustomerKey
 	}
 	return reader, nil
 }
 
-func newCacheEncryptMetadata(bucket, object string, metadata map[string]string) ([]byte, error) {
+func newCacheEncryptMetadata(ctx context.Context, bucket, object string, metadata map[string]string) ([]byte, error) {
 	var sealedKey crypto.SealedKey
 	if globalCacheKMS == nil {
 		return nil, errKMSNotConfigured
 	}
-	key, err := globalCacheKMS.GenerateKey("", kms.Context{bucket: pathJoin(bucket, object)})
+	key, err := globalCacheKMS.GenerateKey(ctx, "", kms.Context{bucket: pathJoin(bucket, object)})
 	if err != nil {
 		return nil, err
 	}
@@ -891,7 +891,7 @@ func (c *diskCache) put(ctx context.Context, bucket, object string, data io.Read
 	reader := data
 	actualSize := uint64(size)
 	if globalCacheKMS != nil {
-		reader, err = newCacheEncryptReader(data, bucket, object, metadata)
+		reader, err = newCacheEncryptReader(ctx, data, bucket, object, metadata)
 		if err != nil {
 			removeAll(cachePath)
 			return oi, err
@@ -948,7 +948,7 @@ func (c *diskCache) putRange(ctx context.Context, bucket, object string, data io
 	// objSize is the actual size of object (with encryption overhead if any)
 	objSize := uint64(size)
 	if globalCacheKMS != nil {
-		reader, err = newCacheEncryptReader(data, bucket, object, metadata)
+		reader, err = newCacheEncryptReader(ctx, data, bucket, object, metadata)
 		if err != nil {
 			return err
 		}
@@ -1275,12 +1275,12 @@ func (c *diskCache) NewMultipartUpload(ctx context.Context, bucket, object, uID
 
 	cachePath := getMultipartCacheSHADir(c.dir, bucket, object)
 	uploadIDDir := path.Join(cachePath, uploadID)
-	if err := os.MkdirAll(uploadIDDir, 0o777); err != nil {
+	if err := mkdirAll(uploadIDDir, 0o777); err != nil {
 		return uploadID, err
 	}
 	metaPath := pathJoin(uploadIDDir, cacheMetaJSONFile)
 
-	f, err := os.OpenFile(metaPath, os.O_RDWR|os.O_CREATE, 0o666)
+	f, err := OpenFile(metaPath, os.O_RDWR|os.O_CREATE|writeMode, 0o666)
 	if err != nil {
 		return uploadID, err
 	}
@@ -1301,7 +1301,7 @@ func (c *diskCache) NewMultipartUpload(ctx context.Context, bucket, object, uID
 	m.Stat.ModTime = UTCNow()
 	if globalCacheKMS != nil {
 		m.Meta[ReservedMetadataPrefix+"Encrypted-Multipart"] = ""
-		if _, err := newCacheEncryptMetadata(bucket, object, m.Meta); err != nil {
+		if _, err := newCacheEncryptMetadata(ctx, bucket, object, m.Meta); err != nil {
 			return uploadID, err
 		}
 	}
@@ -1386,7 +1386,7 @@ func (c *diskCache) SavePartMetadata(ctx context.Context, bucket, object, upload
 	defer uploadLock.Unlock(ulkctx.Cancel)
 
 	metaPath := pathJoin(uploadDir, cacheMetaJSONFile)
-	f, err := os.OpenFile(metaPath, os.O_RDWR, 0o666)
+	f, err := OpenFile(metaPath, os.O_RDWR|writeMode, 0o666)
 	if err != nil {
 		return err
 	}
@@ -1454,7 +1454,7 @@ func newCachePartEncryptReader(ctx context.Context, bucket, object string, partI
 		return nil, err
 	}
 
-	reader, err := sio.EncryptReader(content, sio.Config{Key: partEncryptionKey[:], MinVersion: sio.Version20, CipherSuites: fips.CipherSuitesDARE()})
+	reader, err := sio.EncryptReader(content, sio.Config{Key: partEncryptionKey[:], MinVersion: sio.Version20, CipherSuites: fips.DARECiphers()})
 	if err != nil {
 		return nil, crypto.ErrInvalidCustomerKey
 	}
@@ -1465,7 +1465,7 @@ func newCachePartEncryptReader(ctx context.Context, bucket, object string, partI
 func (c *diskCache) uploadIDExists(bucket, object, uploadID string) (err error) {
 	mpartCachePath := getMultipartCacheSHADir(c.dir, bucket, object)
 	uploadIDDir := path.Join(mpartCachePath, uploadID)
-	if _, err := os.Stat(uploadIDDir); err != nil {
+	if _, err := Stat(uploadIDDir); err != nil {
 		return err
 	}
 	return nil
@@ -1564,7 +1564,7 @@ func (c *diskCache) CompleteMultipartUpload(ctx context.Context, bucket, object,
 	uploadMeta.Hits++
 	metaPath := pathJoin(uploadIDDir, cacheMetaJSONFile)
 
-	f, err := os.OpenFile(metaPath, os.O_RDWR|os.O_CREATE, 0o666)
+	f, err := OpenFile(metaPath, os.O_RDWR|os.O_CREATE|writeMode, 0o666)
 	if err != nil {
 		return oi, err
 	}
@@ -1634,7 +1634,7 @@ func (c *diskCache) cleanupStaleUploads(ctx context.Context) {
 			readDirFn(pathJoin(c.dir, minioMetaBucket, cacheMultipartDir), func(shaDir string, typ os.FileMode) error {
 				return readDirFn(pathJoin(c.dir, minioMetaBucket, cacheMultipartDir, shaDir), func(uploadIDDir string, typ os.FileMode) error {
 					uploadIDPath := pathJoin(c.dir, minioMetaBucket, cacheMultipartDir, shaDir, uploadIDDir)
-					fi, err := os.Stat(uploadIDPath)
+					fi, err := Stat(uploadIDPath)
 					if err != nil {
 						return nil
 					}
@@ -1649,8 +1649,8 @@ func (c *diskCache) cleanupStaleUploads(ctx context.Context) {
 			readDirFn(pathJoin(c.dir, minioMetaBucket, cacheWritebackDir), func(shaDir string, typ os.FileMode) error {
 				wbdir := pathJoin(c.dir, minioMetaBucket, cacheWritebackDir, shaDir)
 				cachedir := pathJoin(c.dir, shaDir)
-				if _, err := os.Stat(cachedir); os.IsNotExist(err) {
-					fi, err := os.Stat(wbdir)
+				if _, err := Stat(cachedir); os.IsNotExist(err) {
+					fi, err := Stat(wbdir)
 					if err != nil {
 						return nil
 					}

cmd/disk-cache.go

@@ -607,12 +607,12 @@ func newCache(config cache.Config) ([]*diskCache, bool, error) {
 				warningMsg = fmt.Sprintf("Invalid cache dir %s err : %s", dir, err.Error())
 			}
 			if rootDsk {
-				warningMsg = fmt.Sprintf("cache dir cannot be part of root disk: %s", dir)
+				warningMsg = fmt.Sprintf("cache dir cannot be part of root drive: %s", dir)
 			}
 		}
 
 		if err := checkAtimeSupport(dir); err != nil {
-			return nil, false, fmt.Errorf("Atime support required for disk caching, atime check failed with %w", err)
+			return nil, false, fmt.Errorf("Atime support required for drive caching, atime check failed with %w", err)
 		}
 
 		cache, err := newDiskCache(ctx, dir, config)
@@ -622,7 +622,7 @@ func newCache(config cache.Config) ([]*diskCache, bool, error) {
 		caches = append(caches, cache)
 	}
 	if warningMsg != "" {
-		logger.Info(color.Yellow(fmt.Sprintf("WARNING: Usage of root disk for disk caching is deprecated: %s", warningMsg)))
+		logger.Info(color.Yellow(fmt.Sprintf("WARNING: Usage of root drive for drive caching is deprecated: %s", warningMsg)))
 	}
 	return caches, migrating, nil
 }

cmd/dummy-handlers.go

@@ -52,7 +52,7 @@ func (api objectAPIHandlers) GetBucketWebsiteHandler(w http.ResponseWriter, r *h
 	}
 
 	// Validate if bucket exists, before proceeding further...
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -84,7 +84,7 @@ func (api objectAPIHandlers) GetBucketAccelerateHandler(w http.ResponseWriter, r
 	}
 
 	// Validate if bucket exists, before proceeding further...
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -117,7 +117,7 @@ func (api objectAPIHandlers) GetBucketRequestPaymentHandler(w http.ResponseWrite
 	}
 
 	// Validate if bucket exists, before proceeding further...
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -151,7 +151,7 @@ func (api objectAPIHandlers) GetBucketLoggingHandler(w http.ResponseWriter, r *h
 	}
 
 	// Validate if bucket exists, before proceeding further...
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -189,7 +189,7 @@ func (api objectAPIHandlers) GetBucketCorsHandler(w http.ResponseWriter, r *http
 	}
 
 	// Validate if bucket exists, before proceeding further...
-	_, err := objAPI.GetBucketInfo(ctx, bucket)
+	_, err := objAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return

cmd/dynamic-timeouts.go

@@ -34,11 +34,24 @@ const (
 
 // timeouts that are dynamically adapted based on actual usage results
 type dynamicTimeout struct {
-	timeout int64
-	minimum int64
-	entries int64
-	log     [dynamicTimeoutLogSize]time.Duration
-	mutex   sync.Mutex
+	timeout       int64
+	minimum       int64
+	entries       int64
+	log           [dynamicTimeoutLogSize]time.Duration
+	mutex         sync.Mutex
+	retryInterval time.Duration
+}
+
+type dynamicTimeoutOpts struct {
+	timeout       time.Duration
+	minimum       time.Duration
+	retryInterval time.Duration
+}
+
+func newDynamicTimeoutWithOpts(opts dynamicTimeoutOpts) *dynamicTimeout {
+	dt := newDynamicTimeout(opts.timeout, opts.minimum)
+	dt.retryInterval = opts.retryInterval
+	return dt
 }
 
 // newDynamicTimeout returns a new dynamic timeout initialized with timeout value
@@ -57,6 +70,10 @@ func (dt *dynamicTimeout) Timeout() time.Duration {
 	return time.Duration(atomic.LoadInt64(&dt.timeout))
 }
 
+func (dt *dynamicTimeout) RetryInterval() time.Duration {
+	return dt.retryInterval
+}
+
 // LogSuccess logs the duration of a successful action that
 // did not hit the timeout
 func (dt *dynamicTimeout) LogSuccess(duration time.Duration) {

cmd/encryption-v1.go

@@ -22,7 +22,6 @@ import (
 	"context"
 	"crypto/hmac"
 	"crypto/rand"
-	"crypto/sha256"
 	"crypto/subtle"
 	"encoding/binary"
 	"encoding/hex"
@@ -38,6 +37,7 @@ import (
 	"github.com/minio/minio/internal/crypto"
 	"github.com/minio/minio/internal/etag"
 	"github.com/minio/minio/internal/fips"
+	"github.com/minio/minio/internal/hash/sha256"
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/kms"
 	"github.com/minio/minio/internal/logger"
@@ -82,7 +82,7 @@ func (o *MultipartInfo) KMSKeyID() string { return kmsKeyIDFromMetadata(o.UserDe
 // metadata, if any. It returns an empty ID if no key ID is
 // present.
 func kmsKeyIDFromMetadata(metadata map[string]string) string {
-	const ARNPrefix = "arn:aws:kms:"
+	const ARNPrefix = crypto.ARNPrefix
 	if len(metadata) == 0 {
 		return ""
 	}
@@ -255,7 +255,7 @@ func ParseSSECustomerHeader(header http.Header) (key []byte, err error) {
 }
 
 // This function rotates old to new key.
-func rotateKey(oldKey []byte, newKeyID string, newKey []byte, bucket, object string, metadata map[string]string, ctx kms.Context) error {
+func rotateKey(ctx context.Context, oldKey []byte, newKeyID string, newKey []byte, bucket, object string, metadata map[string]string, cryptoCtx kms.Context) error {
 	kind, _ := crypto.IsEncrypted(metadata)
 	switch kind {
 	case crypto.S3:
@@ -275,7 +275,7 @@ func rotateKey(oldKey []byte, newKeyID string, newKey []byte, bucket, object str
 			return err
 		}
 
-		newKey, err := GlobalKMS.GenerateKey("", kms.Context{bucket: path.Join(bucket, object)})
+		newKey, err := GlobalKMS.GenerateKey(ctx, "", kms.Context{bucket: path.Join(bucket, object)})
 		if err != nil {
 			return err
 		}
@@ -291,8 +291,8 @@ func rotateKey(oldKey []byte, newKeyID string, newKey []byte, bucket, object str
 			return err
 		}
 
-		if len(ctx) == 0 {
-			_, _, _, ctx, err = crypto.S3KMS.ParseMetadata(metadata)
+		if len(cryptoCtx) == 0 {
+			_, _, _, cryptoCtx, err = crypto.S3KMS.ParseMetadata(metadata)
 			if err != nil {
 				return err
 			}
@@ -305,19 +305,19 @@ func rotateKey(oldKey []byte, newKeyID string, newKey []byte, bucket, object str
 		// of the client provided context and add the bucket
 		// key, if not present.
 		kmsCtx := kms.Context{}
-		for k, v := range ctx {
+		for k, v := range cryptoCtx {
 			kmsCtx[k] = v
 		}
 		if _, ok := kmsCtx[bucket]; !ok {
 			kmsCtx[bucket] = path.Join(bucket, object)
 		}
-		newKey, err := GlobalKMS.GenerateKey(newKeyID, kmsCtx)
+		newKey, err := GlobalKMS.GenerateKey(ctx, newKeyID, kmsCtx)
 		if err != nil {
 			return err
 		}
 
 		sealedKey := objectKey.Seal(newKey.Plaintext, crypto.GenerateIV(rand.Reader), crypto.S3KMS.String(), bucket, object)
-		crypto.S3KMS.CreateMetadata(metadata, newKey.KeyID, newKey.Ciphertext, sealedKey, ctx)
+		crypto.S3KMS.CreateMetadata(metadata, newKey.KeyID, newKey.Ciphertext, sealedKey, cryptoCtx)
 		return nil
 	case crypto.SSEC:
 		sealedKey, err := crypto.SSEC.ParseMetadata(metadata)
@@ -344,14 +344,14 @@ func rotateKey(oldKey []byte, newKeyID string, newKey []byte, bucket, object str
 	}
 }
 
-func newEncryptMetadata(kind crypto.Type, keyID string, key []byte, bucket, object string, metadata map[string]string, ctx kms.Context) (crypto.ObjectKey, error) {
+func newEncryptMetadata(ctx context.Context, kind crypto.Type, keyID string, key []byte, bucket, object string, metadata map[string]string, cryptoCtx kms.Context) (crypto.ObjectKey, error) {
 	var sealedKey crypto.SealedKey
 	switch kind {
 	case crypto.S3:
 		if GlobalKMS == nil {
 			return crypto.ObjectKey{}, errKMSNotConfigured
 		}
-		key, err := GlobalKMS.GenerateKey("", kms.Context{bucket: path.Join(bucket, object)})
+		key, err := GlobalKMS.GenerateKey(ctx, "", kms.Context{bucket: path.Join(bucket, object)})
 		if err != nil {
 			return crypto.ObjectKey{}, err
 		}
@@ -372,13 +372,13 @@ func newEncryptMetadata(kind crypto.Type, keyID string, key []byte, bucket, obje
 		// of the client provided context and add the bucket
 		// key, if not present.
 		kmsCtx := kms.Context{}
-		for k, v := range ctx {
+		for k, v := range cryptoCtx {
 			kmsCtx[k] = v
 		}
 		if _, ok := kmsCtx[bucket]; !ok {
 			kmsCtx[bucket] = path.Join(bucket, object)
 		}
-		key, err := GlobalKMS.GenerateKey(keyID, kmsCtx)
+		key, err := GlobalKMS.GenerateKey(ctx, keyID, kmsCtx)
 		if err != nil {
 			if errors.Is(err, kes.ErrKeyNotFound) {
 				return crypto.ObjectKey{}, errKMSKeyNotFound
@@ -388,7 +388,7 @@ func newEncryptMetadata(kind crypto.Type, keyID string, key []byte, bucket, obje
 
 		objectKey := crypto.GenerateKey(key.Plaintext, rand.Reader)
 		sealedKey = objectKey.Seal(key.Plaintext, crypto.GenerateIV(rand.Reader), crypto.S3KMS.String(), bucket, object)
-		crypto.S3KMS.CreateMetadata(metadata, key.KeyID, key.Ciphertext, sealedKey, ctx)
+		crypto.S3KMS.CreateMetadata(metadata, key.KeyID, key.Ciphertext, sealedKey, cryptoCtx)
 		return objectKey, nil
 	case crypto.SSEC:
 		objectKey := crypto.GenerateKey(key, rand.Reader)
@@ -400,13 +400,13 @@ func newEncryptMetadata(kind crypto.Type, keyID string, key []byte, bucket, obje
 	}
 }
 
-func newEncryptReader(content io.Reader, kind crypto.Type, keyID string, key []byte, bucket, object string, metadata map[string]string, ctx kms.Context) (io.Reader, crypto.ObjectKey, error) {
-	objectEncryptionKey, err := newEncryptMetadata(kind, keyID, key, bucket, object, metadata, ctx)
+func newEncryptReader(ctx context.Context, content io.Reader, kind crypto.Type, keyID string, key []byte, bucket, object string, metadata map[string]string, cryptoCtx kms.Context) (io.Reader, crypto.ObjectKey, error) {
+	objectEncryptionKey, err := newEncryptMetadata(ctx, kind, keyID, key, bucket, object, metadata, cryptoCtx)
 	if err != nil {
 		return nil, crypto.ObjectKey{}, err
 	}
 
-	reader, err := sio.EncryptReader(content, sio.Config{Key: objectEncryptionKey[:], MinVersion: sio.Version20, CipherSuites: fips.CipherSuitesDARE()})
+	reader, err := sio.EncryptReader(content, sio.Config{Key: objectEncryptionKey[:], MinVersion: sio.Version20, CipherSuites: fips.DARECiphers()})
 	if err != nil {
 		return nil, crypto.ObjectKey{}, crypto.ErrInvalidCustomerKey
 	}
@@ -418,9 +418,9 @@ func newEncryptReader(content io.Reader, kind crypto.Type, keyID string, key []b
 // SSE-S3
 func setEncryptionMetadata(r *http.Request, bucket, object string, metadata map[string]string) (err error) {
 	var (
-		key   []byte
-		keyID string
-		ctx   kms.Context
+		key    []byte
+		keyID  string
+		kmsCtx kms.Context
 	)
 	kind, _ := crypto.IsRequested(r.Header)
 	switch kind {
@@ -430,12 +430,12 @@ func setEncryptionMetadata(r *http.Request, bucket, object string, metadata map[
 			return err
 		}
 	case crypto.S3KMS:
-		keyID, ctx, err = crypto.S3KMS.ParseHTTP(r.Header)
+		keyID, kmsCtx, err = crypto.S3KMS.ParseHTTP(r.Header)
 		if err != nil {
 			return err
 		}
 	}
-	_, err = newEncryptMetadata(kind, keyID, key, bucket, object, metadata, ctx)
+	_, err = newEncryptMetadata(r.Context(), kind, keyID, key, bucket, object, metadata, kmsCtx)
 	return
 }
 
@@ -468,7 +468,7 @@ func EncryptRequest(content io.Reader, r *http.Request, bucket, object string, m
 			return nil, crypto.ObjectKey{}, err
 		}
 	}
-	return newEncryptReader(content, kind, keyID, key, bucket, object, metadata, ctx)
+	return newEncryptReader(r.Context(), content, kind, keyID, key, bucket, object, metadata, ctx)
 }
 
 func decryptObjectInfo(key []byte, bucket, object string, metadata map[string]string) ([]byte, error) {
@@ -553,7 +553,7 @@ func newDecryptReaderWithObjectKey(client io.Reader, objectEncryptionKey []byte,
 	reader, err := sio.DecryptReader(client, sio.Config{
 		Key:            objectEncryptionKey,
 		SequenceNumber: seqNumber,
-		CipherSuites:   fips.CipherSuitesDARE(),
+		CipherSuites:   fips.DARECiphers(),
 	})
 	if err != nil {
 		return nil, crypto.ErrInvalidCustomerKey

cmd/endpoint-ellipses.go

@@ -42,7 +42,7 @@ type endpointSet struct {
 
 // Supported set sizes this is used to find the optimal
 // single set size.
-var setSizes = []uint64{4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}
+var setSizes = []uint64{2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}
 
 // getDivisibleSize - returns a greatest common divisor of
 // all the ellipses sizes.
@@ -156,7 +156,7 @@ func getSetIndexes(args []string, totalSizes []uint64, customSetDriveCount uint6
 
 	setCounts := possibleSetCounts(commonSize)
 	if len(setCounts) == 0 {
-		msg := fmt.Sprintf("Incorrect number of endpoints provided %s, number of disks %d is not divisible by any supported erasure set sizes %d", args, commonSize, setSizes)
+		msg := fmt.Sprintf("Incorrect number of endpoints provided %s, number of drives %d is not divisible by any supported erasure set sizes %d", args, commonSize, setSizes)
 		return nil, config.ErrInvalidNumberOfErasureEndpoints(nil).Msg(msg)
 	}
 
@@ -183,7 +183,7 @@ func getSetIndexes(args []string, totalSizes []uint64, customSetDriveCount uint6
 		setCounts = possibleSetCountsWithSymmetry(setCounts, argPatterns)
 
 		if len(setCounts) == 0 {
-			msg := fmt.Sprintf("No symmetric distribution detected with input endpoints provided %s, disks %d cannot be spread symmetrically by any supported erasure set sizes %d", args, commonSize, setSizes)
+			msg := fmt.Sprintf("No symmetric distribution detected with input endpoints provided %s, drives %d cannot be spread symmetrically by any supported erasure set sizes %d", args, commonSize, setSizes)
 			return nil, config.ErrInvalidNumberOfErasureEndpoints(nil).Msg(msg)
 		}
 
@@ -193,7 +193,7 @@ func getSetIndexes(args []string, totalSizes []uint64, customSetDriveCount uint6
 
 	// Check whether setSize is with the supported range.
 	if !isValidSetSize(setSize) {
-		msg := fmt.Sprintf("Incorrect number of endpoints provided %s, number of disks %d is not divisible by any supported erasure set sizes %d", args, commonSize, setSizes)
+		msg := fmt.Sprintf("Incorrect number of endpoints provided %s, number of drives %d is not divisible by any supported erasure set sizes %d", args, commonSize, setSizes)
 		return nil, config.ErrInvalidNumberOfErasureEndpoints(nil).Msg(msg)
 	}
 

cmd/endpoint-ellipses_test.go

@@ -201,18 +201,6 @@ func TestGetSetIndexes(t *testing.T) {
 		success    bool
 	}{
 		// Invalid inputs.
-		{
-			[]string{"data{1...3}"},
-			[]uint64{3},
-			nil,
-			false,
-		},
-		{
-			[]string{"data/controller1/export{1...2}, data/controller2/export{1...4}, data/controller3/export{1...8}"},
-			[]uint64{2, 4, 8},
-			nil,
-			false,
-		},
 		{
 			[]string{"data{1...17}/export{1...52}"},
 			[]uint64{14144},
@@ -220,6 +208,18 @@ func TestGetSetIndexes(t *testing.T) {
 			false,
 		},
 		// Valid inputs.
+		{
+			[]string{"data{1...3}"},
+			[]uint64{3},
+			[][]uint64{{3}},
+			true,
+		},
+		{
+			[]string{"data/controller1/export{1...2}, data/controller2/export{1...4}, data/controller3/export{1...8}"},
+			[]uint64{2, 4, 8},
+			[][]uint64{{2}, {2, 2}, {2, 2, 2, 2}},
+			true,
+		},
 		{
 			[]string{"data{1...27}"},
 			[]uint64{27},

cmd/endpoint.go

@@ -292,6 +292,19 @@ func (l EndpointServerPools) LocalDisksPaths() []string {
 	return disks
 }
 
+// NLocalDisksPathsPerPool returns the disk paths of the local disks per pool
+func (l EndpointServerPools) NLocalDisksPathsPerPool() []int {
+	localDisksCount := make([]int, len(l))
+	for i, ep := range l {
+		for _, endpoint := range ep.Endpoints {
+			if endpoint.IsLocal {
+				localDisksCount[i]++
+			}
+		}
+	}
+	return localDisksCount
+}
+
 // FirstLocal returns true if the first endpoint is local.
 func (l EndpointServerPools) FirstLocal() bool {
 	return l[0].Endpoints[0].IsLocal
@@ -302,7 +315,7 @@ func (l EndpointServerPools) HTTPS() bool {
 	return l[0].Endpoints.HTTPS()
 }
 
-// NEndpoints - returns all nodes count
+// NEndpoints - returns number of endpoints
 func (l EndpointServerPools) NEndpoints() (count int) {
 	for _, ep := range l {
 		count += len(ep.Endpoints)

cmd/erasure-bucket.go

@@ -35,7 +35,7 @@ var bucketMetadataOpIgnoredErrs = append(bucketOpIgnoredErrs, errVolumeNotFound)
 // Bucket operations
 
 // MakeBucket - make a bucket.
-func (er erasureObjects) MakeBucketWithLocation(ctx context.Context, bucket string, opts BucketOptions) error {
+func (er erasureObjects) MakeBucketWithLocation(ctx context.Context, bucket string, opts MakeBucketOptions) error {
 	defer NSUpdated(bucket, slashSeparator)
 
 	// Verify if bucket is valid.
@@ -94,7 +94,7 @@ func undoDeleteBucket(storageDisks []StorageAPI, bucket string) {
 }
 
 // getBucketInfo - returns the BucketInfo from one of the load balanced disks.
-func (er erasureObjects) getBucketInfo(ctx context.Context, bucketName string) (bucketInfo BucketInfo, err error) {
+func (er erasureObjects) getBucketInfo(ctx context.Context, bucketName string, opts BucketOptions) (bucketInfo BucketInfo, err error) {
 	storageDisks := er.getDisks()
 
 	g := errgroup.WithNErrs(len(storageDisks))
@@ -108,9 +108,17 @@ func (er erasureObjects) getBucketInfo(ctx context.Context, bucketName string) (
 			}
 			volInfo, err := storageDisks[index].StatVol(ctx, bucketName)
 			if err != nil {
+				if opts.Deleted {
+					dvi, derr := storageDisks[index].StatVol(ctx, pathJoin(minioMetaBucket, bucketMetaPrefix, deletedBucketsPrefix, bucketName))
+					if derr != nil {
+						return err
+					}
+					bucketsInfo[index] = BucketInfo{Name: bucketName, Deleted: dvi.Created}
+					return nil
+				}
 				return err
 			}
-			bucketsInfo[index] = BucketInfo(volInfo)
+			bucketsInfo[index] = BucketInfo{Name: volInfo.Name, Created: volInfo.Created}
 			return nil
 		}, index)
 	}
@@ -131,8 +139,8 @@ func (er erasureObjects) getBucketInfo(ctx context.Context, bucketName string) (
 }
 
 // GetBucketInfo - returns BucketInfo for a bucket.
-func (er erasureObjects) GetBucketInfo(ctx context.Context, bucket string) (bi BucketInfo, e error) {
-	bucketInfo, err := er.getBucketInfo(ctx, bucket)
+func (er erasureObjects) GetBucketInfo(ctx context.Context, bucket string, opts BucketOptions) (bi BucketInfo, e error) {
+	bucketInfo, err := er.getBucketInfo(ctx, bucket, opts)
 	if err != nil {
 		return bi, toObjectErr(err, bucket)
 	}
@@ -191,12 +199,58 @@ func (er erasureObjects) DeleteBucket(ctx context.Context, bucket string, opts D
 		if purgedDangling {
 			err = nil
 		}
-
+		if opts.SRDeleteOp == MarkDelete {
+			er.markDelete(ctx, minioMetaBucket, pathJoin(bucketMetaPrefix, deletedBucketsPrefix, bucket))
+		}
 	}
 
 	return toObjectErr(err, bucket)
 }
 
+// markDelete creates a vol entry in .minio.sys/buckets/.deleted until site replication
+// syncs the delete to peers
+func (er erasureObjects) markDelete(ctx context.Context, bucket, prefix string) error {
+	storageDisks := er.getDisks()
+	g := errgroup.WithNErrs(len(storageDisks))
+	// Make a volume entry on all underlying storage disks.
+	for index := range storageDisks {
+		index := index
+		if storageDisks[index] == nil {
+			continue
+		}
+		g.Go(func() error {
+			if err := storageDisks[index].MakeVol(ctx, pathJoin(bucket, prefix)); err != nil {
+				if errors.Is(err, errVolumeExists) {
+					return nil
+				}
+				return err
+			}
+			return nil
+		}, index)
+	}
+	err := reduceWriteQuorumErrs(ctx, g.Wait(), bucketOpIgnoredErrs, er.defaultWQuorum())
+	return toObjectErr(err, bucket)
+}
+
+// purgeDelete deletes vol entry in .minio.sys/buckets/.deleted after site replication
+// syncs the delete to peers OR on a new MakeBucket call.
+func (er erasureObjects) purgeDelete(ctx context.Context, bucket, prefix string) error {
+	storageDisks := er.getDisks()
+	g := errgroup.WithNErrs(len(storageDisks))
+	// Make a volume entry on all underlying storage disks.
+	for index := range storageDisks {
+		index := index
+		g.Go(func() error {
+			if storageDisks[index] != nil {
+				return storageDisks[index].DeleteVol(ctx, pathJoin(bucket, prefix), true)
+			}
+			return errDiskNotFound
+		}, index)
+	}
+	err := reduceWriteQuorumErrs(ctx, g.Wait(), bucketOpIgnoredErrs, er.defaultWQuorum())
+	return toObjectErr(err, bucket)
+}
+
 // IsNotificationSupported returns whether bucket notification is applicable for this layer.
 func (er erasureObjects) IsNotificationSupported() bool {
 	return true

cmd/erasure-common.go

@@ -19,7 +19,11 @@ package cmd
 
 import (
 	"context"
+	"fmt"
 	"sync"
+
+	"github.com/minio/minio/internal/logger"
+	"github.com/minio/minio/internal/sync/errgroup"
 )
 
 func (er erasureObjects) getOnlineDisks() (newDisks []StorageAPI) {
@@ -119,3 +123,84 @@ func (er erasureObjects) getLoadBalancedDisks(optimized bool) []StorageAPI {
 	// Return disks which have maximum disk usage common.
 	return newDisks[max]
 }
+
+// readMultipleFiles Reads raw data from all specified files from all disks.
+func readMultipleFiles(ctx context.Context, disks []StorageAPI, req ReadMultipleReq, readQuorum int) ([]ReadMultipleResp, error) {
+	resps := make([]chan ReadMultipleResp, len(disks))
+	for i := range resps {
+		resps[i] = make(chan ReadMultipleResp, len(req.Files))
+	}
+	g := errgroup.WithNErrs(len(disks))
+	// Read files in parallel across disks.
+	for index := range disks {
+		index := index
+		g.Go(func() (err error) {
+			if disks[index] == nil {
+				return errDiskNotFound
+			}
+			return disks[index].ReadMultiple(ctx, req, resps[index])
+		}, index)
+	}
+
+	dataArray := make([]ReadMultipleResp, 0, len(req.Files))
+	// Merge results. They should come in order from each.
+	for _, wantFile := range req.Files {
+		quorum := 0
+		toAdd := ReadMultipleResp{
+			Bucket: req.Bucket,
+			Prefix: req.Prefix,
+			File:   wantFile,
+		}
+		for i := range resps {
+			if disks[i] == nil {
+				continue
+			}
+			select {
+			case <-ctx.Done():
+			case gotFile, ok := <-resps[i]:
+				if !ok {
+					continue
+				}
+				if gotFile.Error != "" || !gotFile.Exists {
+					continue
+				}
+				if gotFile.File != wantFile || gotFile.Bucket != req.Bucket || gotFile.Prefix != req.Prefix {
+					continue
+				}
+				quorum++
+				if toAdd.Modtime.After(gotFile.Modtime) || len(gotFile.Data) < len(toAdd.Data) {
+					// Pick latest, or largest to avoid possible truncated entries.
+					continue
+				}
+				toAdd = gotFile
+			}
+		}
+		if quorum < readQuorum {
+			toAdd.Exists = false
+			toAdd.Error = errErasureReadQuorum.Error()
+			toAdd.Data = nil
+		}
+		dataArray = append(dataArray, toAdd)
+	}
+
+	errs := g.Wait()
+	for index, err := range errs {
+		if err == nil {
+			continue
+		}
+		if !IsErr(err, []error{
+			errFileNotFound,
+			errVolumeNotFound,
+			errFileVersionNotFound,
+			errDiskNotFound,
+			errUnformattedDisk,
+		}...) {
+			logger.LogOnceIf(ctx, fmt.Errorf("Drive %s, path (%s/%s) returned an error (%w)",
+				disks[index], req.Bucket, req.Prefix, err),
+				disks[index].String())
+		}
+	}
+
+	// Return all the metadata.
+	return dataArray, nil
+}

cmd/erasure-decode_test.go

@@ -85,19 +85,17 @@ var erasureDecodeTests = []struct {
 
 func TestErasureDecode(t *testing.T) {
 	for i, test := range erasureDecodeTests {
-		setup, err := newErasureTestSetup(test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
+		setup, err := newErasureTestSetup(t, test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
 		if err != nil {
 			t.Fatalf("Test %d: failed to create test setup: %v", i, err)
 		}
 		erasure, err := NewErasure(context.Background(), test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
 		if err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create ErasureStorage: %v", i, err)
 		}
 		disks := setup.disks
 		data := make([]byte, test.data)
 		if _, err = io.ReadFull(crand.Reader, data); err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to generate random test data: %v", i, err)
 		}
 
@@ -113,11 +111,9 @@ func TestErasureDecode(t *testing.T) {
 		n, err := erasure.Encode(context.Background(), bytes.NewReader(data), writers, buffer, erasure.dataBlocks+1)
 		closeBitrotWriters(writers)
 		if err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create erasure test file: %v", i, err)
 		}
 		if n != test.data {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create erasure test file", i)
 		}
 		for i, w := range writers {
@@ -195,7 +191,6 @@ func TestErasureDecode(t *testing.T) {
 				}
 			}
 		}
-		setup.Remove()
 	}
 }
 
@@ -210,12 +205,11 @@ func TestErasureDecodeRandomOffsetLength(t *testing.T) {
 	dataBlocks := 7
 	parityBlocks := 7
 	blockSize := int64(1 * humanize.MiByte)
-	setup, err := newErasureTestSetup(dataBlocks, parityBlocks, blockSize)
+	setup, err := newErasureTestSetup(t, dataBlocks, parityBlocks, blockSize)
 	if err != nil {
 		t.Error(err)
 		return
 	}
-	defer setup.Remove()
 	disks := setup.disks
 	erasure, err := NewErasure(context.Background(), dataBlocks, parityBlocks, blockSize)
 	if err != nil {
@@ -288,11 +282,10 @@ func TestErasureDecodeRandomOffsetLength(t *testing.T) {
 // Benchmarks
 
 func benchmarkErasureDecode(data, parity, dataDown, parityDown int, size int64, b *testing.B) {
-	setup, err := newErasureTestSetup(data, parity, blockSizeV2)
+	setup, err := newErasureTestSetup(b, data, parity, blockSizeV2)
 	if err != nil {
 		b.Fatalf("failed to create test setup: %v", err)
 	}
-	defer setup.Remove()
 	disks := setup.disks
 	erasure, err := NewErasure(context.Background(), data, parity, blockSizeV2)
 	if err != nil {

cmd/erasure-encode_test.go

@@ -87,21 +87,19 @@ var erasureEncodeTests = []struct {
 
 func TestErasureEncode(t *testing.T) {
 	for i, test := range erasureEncodeTests {
-		setup, err := newErasureTestSetup(test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
+		setup, err := newErasureTestSetup(t, test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
 		if err != nil {
 			t.Fatalf("Test %d: failed to create test setup: %v", i, err)
 		}
 		disks := setup.disks
 		erasure, err := NewErasure(context.Background(), test.dataBlocks, test.onDisks-test.dataBlocks, test.blocksize)
 		if err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create ErasureStorage: %v", i, err)
 		}
 		buffer := make([]byte, test.blocksize, 2*test.blocksize)
 
 		data := make([]byte, test.data)
 		if _, err = io.ReadFull(rand.Reader, data); err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to generate random test data: %v", i, err)
 		}
 		writers := make([]io.Writer, len(disks))
@@ -160,18 +158,16 @@ func TestErasureEncode(t *testing.T) {
 				}
 			}
 		}
-		setup.Remove()
 	}
 }
 
 // Benchmarks
 
 func benchmarkErasureEncode(data, parity, dataDown, parityDown int, size int64, b *testing.B) {
-	setup, err := newErasureTestSetup(data, parity, blockSizeV2)
+	setup, err := newErasureTestSetup(b, data, parity, blockSizeV2)
 	if err != nil {
 		b.Fatalf("failed to create test setup: %v", err)
 	}
-	defer setup.Remove()
 	erasure, err := NewErasure(context.Background(), data, parity, blockSizeV2)
 	if err != nil {
 		b.Fatalf("failed to create ErasureStorage: %v", err)
@@ -196,7 +192,10 @@ func benchmarkErasureEncode(data, parity, dataDown, parityDown int, size int64,
 			if disk == OfflineDisk {
 				continue
 			}
-			disk.Delete(context.Background(), "testbucket", "object", false)
+			disk.Delete(context.Background(), "testbucket", "object", DeleteOptions{
+				Recursive: false,
+				Force:     false,
+			})
 			writers[i] = newBitrotWriter(disk, "testbucket", "object", erasure.ShardFileSize(size), DefaultBitrotAlgorithm, erasure.ShardSize())
 		}
 		_, err := erasure.Encode(context.Background(), bytes.NewReader(content), writers, buffer, erasure.dataBlocks+1)

cmd/erasure-errors.go

@@ -20,10 +20,10 @@ package cmd
 import "errors"
 
 // errErasureReadQuorum - did not meet read quorum.
-var errErasureReadQuorum = errors.New("Read failed. Insufficient number of disks online")
+var errErasureReadQuorum = errors.New("Read failed. Insufficient number of drives online")
 
 // errErasureWriteQuorum - did not meet write quorum.
-var errErasureWriteQuorum = errors.New("Write failed. Insufficient number of disks online")
+var errErasureWriteQuorum = errors.New("Write failed. Insufficient number of drives online")
 
 // errNoHealRequired - returned when healing is attempted on a previously healed disks.
 var errNoHealRequired = errors.New("No healing is required")

cmd/erasure-heal_test.go

@@ -66,23 +66,21 @@ func TestErasureHeal(t *testing.T) {
 	for i, test := range erasureHealTests {
 		if test.offDisks < test.badStaleDisks {
 			// test case sanity check
-			t.Fatalf("Test %d: Bad test case - number of stale disks cannot be less than number of badstale disks", i)
+			t.Fatalf("Test %d: Bad test case - number of stale drives cannot be less than number of badstale drives", i)
 		}
 
 		// create some test data
-		setup, err := newErasureTestSetup(test.dataBlocks, test.disks-test.dataBlocks, test.blocksize)
+		setup, err := newErasureTestSetup(t, test.dataBlocks, test.disks-test.dataBlocks, test.blocksize)
 		if err != nil {
 			t.Fatalf("Test %d: failed to setup Erasure environment: %v", i, err)
 		}
 		disks := setup.disks
 		erasure, err := NewErasure(context.Background(), test.dataBlocks, test.disks-test.dataBlocks, test.blocksize)
 		if err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create ErasureStorage: %v", i, err)
 		}
 		data := make([]byte, test.size)
 		if _, err = io.ReadFull(rand.Reader, data); err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create random test data: %v", i, err)
 		}
 		buffer := make([]byte, test.blocksize, 2*test.blocksize)
@@ -93,7 +91,6 @@ func TestErasureHeal(t *testing.T) {
 		_, err = erasure.Encode(context.Background(), bytes.NewReader(data), writers, buffer, erasure.dataBlocks+1)
 		closeBitrotWriters(writers)
 		if err != nil {
-			setup.Remove()
 			t.Fatalf("Test %d: failed to create random test data: %v", i, err)
 		}
 
@@ -156,6 +153,5 @@ func TestErasureHeal(t *testing.T) {
 				}
 			}
 		}
-		setup.Remove()
 	}
 }

cmd/erasure-healing-common.go

@@ -180,9 +180,15 @@ func listOnlineDisks(disks []StorageAPI, partsMetadata []FileInfo, errs []error)
 }
 
 // Returns the latest updated FileInfo files and error in case of failure.
-func getLatestFileInfo(ctx context.Context, partsMetadata []FileInfo, errs []error) (FileInfo, error) {
+func getLatestFileInfo(ctx context.Context, partsMetadata []FileInfo, defaultParityCount int, errs []error) (FileInfo, error) {
 	// There should be atleast half correct entries, if not return failure
-	reducedErr := reduceReadQuorumErrs(ctx, errs, objectOpIgnoredErrs, len(partsMetadata)/2)
+	expectedRQuorum := len(partsMetadata) / 2
+	if defaultParityCount == 0 {
+		// if parity count is '0', we expected all entries to be present.
+		expectedRQuorum = len(partsMetadata)
+	}
+
+	reducedErr := reduceReadQuorumErrs(ctx, errs, objectOpIgnoredErrs, expectedRQuorum)
 	if reducedErr != nil {
 		return FileInfo{}, reducedErr
 	}

cmd/erasure-healing-common_test.go

@@ -176,7 +176,7 @@ func TestListOnlineDisks(t *testing.T) {
 	}
 
 	bucket := "bucket"
-	err = obj.MakeBucketWithLocation(ctx, "bucket", BucketOptions{})
+	err = obj.MakeBucketWithLocation(ctx, "bucket", MakeBucketOptions{})
 	if err != nil {
 		t.Fatalf("Failed to make a bucket %v", err)
 	}
@@ -194,7 +194,7 @@ func TestListOnlineDisks(t *testing.T) {
 			}
 
 			partsMetadata, errs := readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
-			fi, err := getLatestFileInfo(ctx, partsMetadata, errs)
+			fi, err := getLatestFileInfo(ctx, partsMetadata, z.serverPools[0].sets[0].defaultParityCount, errs)
 			if err != nil {
 				t.Fatalf("Failed to getLatestFileInfo %v", err)
 			}
@@ -218,7 +218,10 @@ func TestListOnlineDisks(t *testing.T) {
 					// and check if that disk
 					// appears in outDatedDisks.
 					tamperedIndex = index
-					dErr := erasureDisks[index].Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), false)
+					dErr := erasureDisks[index].Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), DeleteOptions{
+						Recursive: false,
+						Force:     false,
+					})
 					if dErr != nil {
 						t.Fatalf("Failed to delete %s - %v", filepath.Join(object, "part.1"), dErr)
 					}
@@ -257,7 +260,7 @@ func TestListOnlineDisks(t *testing.T) {
 
 			if test._tamperBackend != noTamper {
 				if tamperedIndex != -1 && availableDisks[tamperedIndex] != nil {
-					t.Fatalf("disk (%v) with part.1 missing is not a disk with available data",
+					t.Fatalf("Drive (%v) with part.1 missing is not a drive with available data",
 						erasureDisks[tamperedIndex])
 				}
 			}
@@ -348,7 +351,7 @@ func TestListOnlineDisksSmallObjects(t *testing.T) {
 	}
 
 	bucket := "bucket"
-	err = obj.MakeBucketWithLocation(ctx, "bucket", BucketOptions{})
+	err = obj.MakeBucketWithLocation(ctx, "bucket", MakeBucketOptions{})
 	if err != nil {
 		t.Fatalf("Failed to make a bucket %v", err)
 	}
@@ -367,7 +370,7 @@ func TestListOnlineDisksSmallObjects(t *testing.T) {
 			}
 
 			partsMetadata, errs := readAllFileInfo(ctx, erasureDisks, bucket, object, "", true)
-			_, err = getLatestFileInfo(ctx, partsMetadata, errs)
+			_, err = getLatestFileInfo(ctx, partsMetadata, z.serverPools[0].sets[0].defaultParityCount, errs)
 			if err != nil {
 				t.Fatalf("Failed to getLatestFileInfo %v", err)
 			}
@@ -395,7 +398,10 @@ func TestListOnlineDisksSmallObjects(t *testing.T) {
 					// and check if that disk
 					// appears in outDatedDisks.
 					tamperedIndex = index
-					dErr := erasureDisks[index].Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
+					dErr := erasureDisks[index].Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), DeleteOptions{
+						Recursive: false,
+						Force:     false,
+					})
 					if dErr != nil {
 						t.Fatalf("Failed to delete %s - %v", pathJoin(object, xlStorageFormatFile), dErr)
 					}
@@ -423,7 +429,7 @@ func TestListOnlineDisksSmallObjects(t *testing.T) {
 
 			}
 			partsMetadata, errs = readAllFileInfo(ctx, erasureDisks, bucket, object, "", true)
-			fi, err := getLatestFileInfo(ctx, partsMetadata, errs)
+			fi, err := getLatestFileInfo(ctx, partsMetadata, z.serverPools[0].sets[0].defaultParityCount, errs)
 			if !errors.Is(err, errErasureReadQuorum) {
 				t.Fatalf("Failed to getLatestFileInfo, expected %v, got %v", errErasureReadQuorum, err)
 			}
@@ -440,7 +446,7 @@ func TestListOnlineDisksSmallObjects(t *testing.T) {
 
 			if test._tamperBackend != noTamper {
 				if tamperedIndex != -1 && availableDisks[tamperedIndex] != nil {
-					t.Fatalf("disk (%v) with part.1 missing is not a disk with available data",
+					t.Fatalf("Drive (%v) with part.1 missing is not a drive with available data",
 						erasureDisks[tamperedIndex])
 				}
 			}
@@ -466,7 +472,7 @@ func TestDisksWithAllParts(t *testing.T) {
 	z := obj.(*erasureServerPools)
 	s := z.serverPools[0].sets[0]
 	erasureDisks := s.getDisks()
-	err = obj.MakeBucketWithLocation(ctx, "bucket", BucketOptions{})
+	err = obj.MakeBucketWithLocation(ctx, "bucket", MakeBucketOptions{})
 	if err != nil {
 		t.Fatalf("Failed to make a bucket %v", err)
 	}
@@ -489,7 +495,7 @@ func TestDisksWithAllParts(t *testing.T) {
 		t.Fatalf("Failed to read xl meta data %v", err)
 	}
 
-	fi, err := getLatestFileInfo(ctx, partsMetadata, errs)
+	fi, err := getLatestFileInfo(ctx, partsMetadata, s.defaultParityCount, errs)
 	if err != nil {
 		t.Fatalf("Failed to get quorum consistent fileInfo %v", err)
 	}
@@ -500,7 +506,7 @@ func TestDisksWithAllParts(t *testing.T) {
 		errs, fi, bucket, object, madmin.HealDeepScan)
 
 	if len(filteredDisks) != len(erasureDisks) {
-		t.Errorf("Unexpected number of disks: %d", len(filteredDisks))
+		t.Errorf("Unexpected number of drives: %d", len(filteredDisks))
 	}
 
 	for diskIndex, disk := range filteredDisks {
@@ -509,7 +515,7 @@ func TestDisksWithAllParts(t *testing.T) {
 		}
 
 		if disk == nil {
-			t.Errorf("Disk erroneously filtered, diskIndex: %d", diskIndex)
+			t.Errorf("Drive erroneously filtered, driveIndex: %d", diskIndex)
 		}
 	}
 
@@ -522,14 +528,14 @@ func TestDisksWithAllParts(t *testing.T) {
 		errs, fi, bucket, object, madmin.HealDeepScan)
 
 	if len(filteredDisks) != len(erasureDisks) {
-		t.Errorf("Unexpected number of disks: %d", len(filteredDisks))
+		t.Errorf("Unexpected number of drives: %d", len(filteredDisks))
 	}
 	for diskIndex, disk := range filteredDisks {
 		if diskIndex == 0 && disk != nil {
-			t.Errorf("Disk not filtered as expected, disk: %d", diskIndex)
+			t.Errorf("Drive not filtered as expected, drive: %d", diskIndex)
 		}
 		if diskIndex != 0 && disk == nil {
-			t.Errorf("Disk erroneously filtered, diskIndex: %d", diskIndex)
+			t.Errorf("Drive erroneously filtered, driveIndex: %d", diskIndex)
 		}
 	}
 	partsMetadata[0] = partsMetadataBackup // Revert before going to the next test
@@ -543,14 +549,14 @@ func TestDisksWithAllParts(t *testing.T) {
 		errs, fi, bucket, object, madmin.HealDeepScan)
 
 	if len(filteredDisks) != len(erasureDisks) {
-		t.Errorf("Unexpected number of disks: %d", len(filteredDisks))
+		t.Errorf("Unexpected number of drives: %d", len(filteredDisks))
 	}
 	for diskIndex, disk := range filteredDisks {
 		if diskIndex == 1 && disk != nil {
-			t.Errorf("Disk not filtered as expected, disk: %d", diskIndex)
+			t.Errorf("Drive not filtered as expected, drive: %d", diskIndex)
 		}
 		if diskIndex != 1 && disk == nil {
-			t.Errorf("Disk erroneously filtered, diskIndex: %d", diskIndex)
+			t.Errorf("Drive erroneously filtered, driveIndex: %d", diskIndex)
 		}
 	}
 	partsMetadata[1] = partsMetadataBackup // Revert before going to the next test
@@ -580,23 +586,23 @@ func TestDisksWithAllParts(t *testing.T) {
 		errs, fi, bucket, object, madmin.HealDeepScan)
 
 	if len(filteredDisks) != len(erasureDisks) {
-		t.Errorf("Unexpected number of disks: %d", len(filteredDisks))
+		t.Errorf("Unexpected number of drives: %d", len(filteredDisks))
 	}
 
 	for diskIndex, disk := range filteredDisks {
 		if _, ok := diskFailures[diskIndex]; ok {
 			if disk != nil {
-				t.Errorf("Disk not filtered as expected, disk: %d", diskIndex)
+				t.Errorf("Drive not filtered as expected, drive: %d", diskIndex)
 			}
 			if errs[diskIndex] == nil {
-				t.Errorf("Expected error not received, diskIndex: %d", diskIndex)
+				t.Errorf("Expected error not received, driveIndex: %d", diskIndex)
 			}
 		} else {
 			if disk == nil {
-				t.Errorf("Disk erroneously filtered, diskIndex: %d", diskIndex)
+				t.Errorf("Drive erroneously filtered, driveIndex: %d", diskIndex)
 			}
 			if errs[diskIndex] != nil {
-				t.Errorf("Unexpected error, %s, diskIndex: %d", errs[diskIndex], diskIndex)
+				t.Errorf("Unexpected error, %s, driveIndex: %d", errs[diskIndex], diskIndex)
 			}
 
 		}

cmd/erasure-healing.go

@@ -243,7 +243,7 @@ func listAllBuckets(ctx context.Context, storageDisks []StorageAPI, healBuckets
 
 // Only heal on disks where we are sure that healing is needed. We can expand
 // this list as and when we figure out more errors can be added to this list safely.
-func shouldHealObjectOnDisk(erErr, dataErr error, meta FileInfo, latestMeta FileInfo) bool {
+func shouldHealObjectOnDisk(erErr, dataErr error, meta FileInfo, latestMeta FileInfo, doinline bool) bool {
 	switch {
 	case errors.Is(erErr, errFileNotFound) || errors.Is(erErr, errFileVersionNotFound):
 		return true
@@ -256,6 +256,10 @@ func shouldHealObjectOnDisk(erErr, dataErr error, meta FileInfo, latestMeta File
 			// always check first.
 			return true
 		}
+		if doinline {
+			// convert small files to 'inline'
+			return true
+		}
 		if !meta.Deleted && !meta.IsRemote() {
 			// If xl.meta was read fine but there may be problem with the part.N files.
 			if IsErr(dataErr, []error{
@@ -275,10 +279,6 @@ func shouldHealObjectOnDisk(erErr, dataErr error, meta FileInfo, latestMeta File
 
 // Heals an object by re-writing corrupt/missing erasure blocks.
 func (er erasureObjects) healObject(ctx context.Context, bucket string, object string, versionID string, opts madmin.HealOpts) (result madmin.HealResultItem, err error) {
-	if !opts.DryRun {
-		defer NSUpdated(bucket, object)
-	}
-
 	dryRun := opts.DryRun
 	scanMode := opts.ScanMode
 
@@ -306,9 +306,13 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 	// Re-read when we have lock...
 	partsMetadata, errs := readAllFileInfo(ctx, storageDisks, bucket, object, versionID, true)
 	if isAllNotFound(errs) {
+		err := errFileNotFound
+		if versionID != "" {
+			err = errFileVersionNotFound
+		}
 		// Nothing to do, file is already gone.
 		return er.defaultHealResult(FileInfo{}, storageDisks, storageEndpoints,
-			errs, bucket, object, versionID), nil
+			errs, bucket, object, versionID), err
 	}
 
 	readQuorum, _, err := objectQuorumFromMeta(ctx, partsMetadata, errs, er.defaultParityCount)
@@ -327,7 +331,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 	// present, it is as good as object not found.
 	latestMeta, err := pickValidFileInfo(ctx, partsMetadata, modTime, readQuorum)
 	if err != nil {
-		return result, toObjectErr(err, bucket, object, versionID)
+		return result, err
 	}
 
 	// List of disks having all parts as per latest metadata.
@@ -342,6 +346,26 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 	availableDisks, dataErrs, diskMTime := disksWithAllParts(ctx, onlineDisks, partsMetadata,
 		errs, latestMeta, bucket, object, scanMode)
 
+	var erasure Erasure
+	var recreate bool
+	if !latestMeta.Deleted && !latestMeta.IsRemote() {
+		// Initialize erasure coding
+		erasure, err = NewErasure(ctx, latestMeta.Erasure.DataBlocks,
+			latestMeta.Erasure.ParityBlocks, latestMeta.Erasure.BlockSize)
+		if err != nil {
+			return result, err
+		}
+
+		// Is only 'true' if the opts.Recreate is true and
+		// the object shardSize < smallFileThreshold do not
+		// set this to 'true' arbitrarily and must be only
+		// 'true' with caller ask.
+		recreate = (opts.Recreate &&
+			!latestMeta.InlineData() &&
+			len(latestMeta.Parts) == 1 &&
+			erasure.ShardFileSize(latestMeta.Parts[0].ActualSize) < smallFileThreshold)
+	}
+
 	// Loop to find number of disks with valid data, per-drive
 	// data state and a list of outdated disks on which data needs
 	// to be healed.
@@ -368,7 +392,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 			driveState = madmin.DriveStateCorrupt
 		}
 
-		if shouldHealObjectOnDisk(errs[i], dataErrs[i], partsMetadata[i], latestMeta) {
+		if shouldHealObjectOnDisk(errs[i], dataErrs[i], partsMetadata[i], latestMeta, recreate) {
 			outDatedDisks[i] = storageDisks[i]
 			disksToHealCount++
 			result.Before.Drives = append(result.Before.Drives, madmin.HealDriveInfo{
@@ -397,8 +421,12 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 
 	if isAllNotFound(errs) {
 		// File is fully gone, fileInfo is empty.
+		err := errFileNotFound
+		if versionID != "" {
+			err = errFileVersionNotFound
+		}
 		return er.defaultHealResult(FileInfo{}, storageDisks, storageEndpoints, errs,
-			bucket, object, versionID), nil
+			bucket, object, versionID), err
 	}
 
 	// If less than read quorum number of disks have all the parts
@@ -418,9 +446,9 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		return result, nil
 	}
 
-	if !latestMeta.XLV1 && !latestMeta.Deleted && disksToHealCount > latestMeta.Erasure.ParityBlocks {
+	if !latestMeta.XLV1 && !latestMeta.Deleted && !recreate && disksToHealCount > latestMeta.Erasure.ParityBlocks {
 		// When disk to heal count is greater than parity blocks we should simply error out.
-		err := fmt.Errorf("more disks are expected to heal than parity, returned errors: %v -> %s/%s(%s)", errs, bucket, object, versionID)
+		err := fmt.Errorf("more drives are expected to heal than parity, returned errors: %v (dataErrs %v) -> %s/%s(%s)", errs, dataErrs, bucket, object, versionID)
 		logger.LogIf(ctx, err)
 		return er.defaultHealResult(latestMeta, storageDisks, storageEndpoints, errs,
 			bucket, object, versionID), err
@@ -474,25 +502,18 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 	}
 
 	var inlineBuffers []*bytes.Buffer
-	if latestMeta.InlineData() {
-		inlineBuffers = make([]*bytes.Buffer, len(outDatedDisks))
-	}
-
 	if !latestMeta.Deleted && !latestMeta.IsRemote() {
-		// Heal each part. erasureHealFile() will write the healed
-		// part to .minio/tmp/uuid/ which needs to be renamed later to
-		// the final location.
-		erasure, err := NewErasure(ctx, latestMeta.Erasure.DataBlocks,
-			latestMeta.Erasure.ParityBlocks, latestMeta.Erasure.BlockSize)
-		if err != nil {
-			return result, toObjectErr(err, bucket, object)
+		if latestMeta.InlineData() || recreate {
+			inlineBuffers = make([]*bytes.Buffer, len(outDatedDisks))
 		}
 
 		erasureInfo := latestMeta.Erasure
 		for partIndex := 0; partIndex < len(latestMeta.Parts); partIndex++ {
 			partSize := latestMeta.Parts[partIndex].Size
 			partActualSize := latestMeta.Parts[partIndex].ActualSize
+			partModTime := latestMeta.Parts[partIndex].ModTime
 			partNumber := latestMeta.Parts[partIndex].Number
+			partIdx := latestMeta.Parts[partIndex].Index
 			tillOffset := erasure.ShardFileOffset(0, partSize, partSize)
 			readers := make([]io.ReaderAt, len(latestDisks))
 			checksumAlgo := erasureInfo.GetChecksumInfo(partNumber).Algorithm
@@ -519,11 +540,15 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 						tillOffset, DefaultBitrotAlgorithm, erasure.ShardSize())
 				}
 			}
+
+			// Heal each part. erasure.Heal() will write the healed
+			// part to .minio/tmp/uuid/ which needs to be renamed
+			// later to the final location.
 			err = erasure.Heal(ctx, writers, readers, partSize)
 			closeBitrotReaders(readers)
 			closeBitrotWriters(writers)
 			if err != nil {
-				return result, toObjectErr(err, bucket, object)
+				return result, err
 			}
 
 			// outDatedDisks that had write errors should not be
@@ -542,7 +567,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 				}
 
 				partsMetadata[i].DataDir = dstDataDir
-				partsMetadata[i].AddObjectPart(partNumber, "", partSize, partActualSize)
+				partsMetadata[i].AddObjectPart(partNumber, "", partSize, partActualSize, partModTime, partIdx)
 				partsMetadata[i].Erasure.AddChecksumInfo(ChecksumInfo{
 					PartNumber: partNumber,
 					Algorithm:  checksumAlgo,
@@ -550,6 +575,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 				})
 				if len(inlineBuffers) > 0 && inlineBuffers[i] != nil {
 					partsMetadata[i].Data = inlineBuffers[i].Bytes()
+					partsMetadata[i].SetInlineData()
 				} else {
 					partsMetadata[i].Data = nil
 				}
@@ -557,7 +583,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 
 			// If all disks are having errors, we give up.
 			if disksToHealCount == 0 {
-				return result, fmt.Errorf("all disks had write errors, unable to heal %s/%s", bucket, object)
+				return result, fmt.Errorf("all drives had write errors, unable to heal %s/%s", bucket, object)
 			}
 
 		}
@@ -578,11 +604,12 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		// Attempt a rename now from healed data to final location.
 		if err = disk.RenameData(ctx, minioMetaTmpBucket, tmpID, partsMetadata[i], bucket, object); err != nil {
 			logger.LogIf(ctx, err)
-			return result, toObjectErr(err, bucket, object)
+			return result, err
 		}
 
-		// Remove any remaining parts from outdated disks from before transition.
-		if partsMetadata[i].IsRemote() {
+		// - Remove any parts from healed disks after its been inlined.
+		// - Remove any remaining parts from outdated disks from before transition.
+		if recreate || partsMetadata[i].IsRemote() {
 			rmDataDir := partsMetadata[i].DataDir
 			disk.DeleteVol(ctx, pathJoin(bucket, encodeDirObject(object), rmDataDir), true)
 		}
@@ -647,7 +674,10 @@ func (er erasureObjects) healObjectDir(ctx context.Context, bucket, object strin
 				wg.Add(1)
 				go func(index int, disk StorageAPI) {
 					defer wg.Done()
-					_ = disk.Delete(ctx, bucket, object, false)
+					_ = disk.Delete(ctx, bucket, object, DeleteOptions{
+						Recursive: false,
+						Force:     false,
+					})
 				}(index, disk)
 			}
 			wg.Wait()
@@ -674,10 +704,16 @@ func (er erasureObjects) healObjectDir(ctx context.Context, bucket, object strin
 			hr.After.Drives[i] = madmin.HealDriveInfo{Endpoint: drive, State: madmin.DriveStateCorrupt}
 		}
 	}
-	if dryRun || danglingObject || isAllNotFound(errs) {
+	if danglingObject || isAllNotFound(errs) {
 		// Nothing to do, file is already gone.
+		return hr, errFileNotFound
+	}
+
+	if dryRun {
+		// Quit without try to heal the object dir
 		return hr, nil
 	}
+
 	for i, err := range errs {
 		if err == errVolumeNotFound || err == errFileNotFound {
 			// Bucket or prefix/directory not found
@@ -837,8 +873,13 @@ func (er erasureObjects) purgeObjectDangling(ctx context.Context, bucket, object
 		// Dangling object successfully purged, size is '0'
 		m.Size = 0
 	}
+	// Generate file/version not found with default heal result
+	err = errFileNotFound
+	if versionID != "" {
+		err = errFileVersionNotFound
+	}
 	return er.defaultHealResult(m, storageDisks, storageEndpoints,
-		errs, bucket, object, versionID), nil
+		errs, bucket, object, versionID), err
 }
 
 // Object is considered dangling/corrupted if any only
@@ -908,12 +949,6 @@ func isObjectDangling(metaArr []FileInfo, errs []error, dataErrs []error) (valid
 
 // HealObject - heal the given object, automatically deletes the object if stale/corrupted if `remove` is true.
 func (er erasureObjects) HealObject(ctx context.Context, bucket, object, versionID string, opts madmin.HealOpts) (hr madmin.HealResultItem, err error) {
-	defer func() {
-		if isErrObjectNotFound(err) || isErrVersionNotFound(err) {
-			err = nil
-		}
-	}()
-
 	// Create context that also contains information about the object and bucket.
 	// The top level handler might not have this information.
 	reqInfo := logger.GetReqInfo(ctx)
@@ -927,7 +962,8 @@ func (er erasureObjects) HealObject(ctx context.Context, bucket, object, version
 
 	// Healing directories handle it separately.
 	if HasSuffix(object, SlashSeparator) {
-		return er.healObjectDir(healCtx, bucket, object, opts.DryRun, opts.Remove)
+		hr, err := er.healObjectDir(healCtx, bucket, object, opts.DryRun, opts.Remove)
+		return hr, toObjectErr(err, bucket, object)
 	}
 
 	storageDisks := er.getDisks()
@@ -942,9 +978,13 @@ func (er erasureObjects) HealObject(ctx context.Context, bucket, object, version
 	// This allows to quickly check if all is ok or all are missing.
 	_, errs := readAllFileInfo(healCtx, storageDisks, bucket, object, versionID, false)
 	if isAllNotFound(errs) {
+		err := errFileNotFound
+		if versionID != "" {
+			err = errFileVersionNotFound
+		}
 		// Nothing to do, file is already gone.
 		return er.defaultHealResult(FileInfo{}, storageDisks, storageEndpoints,
-			errs, bucket, object, versionID), nil
+			errs, bucket, object, versionID), toObjectErr(err, bucket, object, versionID)
 	}
 
 	// Heal the object.
@@ -955,5 +995,5 @@ func (er erasureObjects) HealObject(ctx context.Context, bucket, object, version
 		opts.ScanMode = madmin.HealDeepScan
 		hr, err = er.healObject(healCtx, bucket, object, versionID, opts)
 	}
-	return hr, err
+	return hr, toObjectErr(err, bucket, object, versionID)
 }