sts: Avoid printing all STS errors (#15065)

Limit printing STS errors to 

- STS internal error
- STS not initialized
- STS upstream error

.github/markdown-lint-cfg.yaml

@@ -0,0 +1,5 @@
+# Config file for markdownlint-cli
+MD033:
+  allowed_elements:
+    - details
+    - summary

.github/workflows/go-cross.yml

@@ -11,19 +11,23 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   build:
     name: Build Tests with Go ${{ matrix.go-version }} on ${{ matrix.os }}
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.17.x, 1.18.x]
         os: [ubuntu-latest]
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@bfdd3570ce990073878bf10f6b2d79082de49492 # v2
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - name: Build on ${{ matrix.os }}
         if: matrix.os == 'ubuntu-latest'
         env:

.github/workflows/go-fips.yml

@@ -0,0 +1,51 @@
+name: FIPS Build Test
+
+on:
+  pull_request:
+    branches:
+    - master
+
+# This ensures that previous jobs for the PR are canceled when the PR is
+# updated.
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    name: Go BoringCrypto ${{ matrix.go-version }} on ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        go-version: [1.17.11b7, 1.18.3b7]
+        os: [ubuntu-latest]
+    steps:
+      - uses: actions/checkout@v2
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v2
+
+      - name: Setup dockerfile for build test
+        run: |
+          echo "FROM us-docker.pkg.dev/google.com/api-project-999119582588/go-boringcrypto/golang:${{ matrix.go-version }}" > Dockerfile.fips.test
+          echo "COPY . /minio" >> Dockerfile.fips.test
+          echo "WORKDIR /minio" >> Dockerfile.fips.test
+          echo "RUN make" >> Dockerfile.fips.test
+
+      - name: Build
+        uses: docker/build-push-action@v3
+        with:
+          context: .
+          file: Dockerfile.fips.test
+          push: false
+          load: true
+          tags: minio/fips-test:latest
+
+      # This should fail if grep returns non-zero exit
+      - name: Test binary
+        run: |
+          docker run --rm minio/fips-test:latest ./minio --version
+          docker run --rm -i minio/fips-test:latest /bin/bash -c 'go tool nm ./minio' | grep -q FIPS

.github/workflows/go-healing.yml

@@ -11,6 +11,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   build:
     name: Go ${{ matrix.go-version }} on ${{ matrix.os }}
@@ -21,9 +24,10 @@ jobs:
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - uses: actions/cache@v2
         with:
           path: |
@@ -37,12 +41,10 @@ jobs:
         env:
           CGO_ENABLED: 0
           GO111MODULE: on
-          MINIO_KMS_KES_CERT_FILE: /home/runner/work/minio/minio/.github/workflows/root.cert
+          MINIO_KMS_SECRET_KEY: "my-minio-key:oyArl7zlPECEduNbB1KXgdzDn2Bdpvvw0l8VO51HQnY="
-          MINIO_KMS_KES_KEY_FILE: /home/runner/work/minio/minio/.github/workflows/root.key
-          MINIO_KMS_KES_ENDPOINT: "https://play.min.io:7373"
-          MINIO_KMS_KES_KEY_NAME: "my-minio-key"
           MINIO_KMS_AUTO_ENCRYPTION: on
         run: |
           sudo sysctl net.ipv6.conf.all.disable_ipv6=0
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make verify-healing
+          make verify-healing-inconsistent-versions

.github/workflows/go-lint.yml

@@ -11,19 +11,23 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   build:
     name: Go ${{ matrix.go-version }} on ${{ matrix.os }}
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.17.x, 1.18.x]
         os: [ubuntu-latest, windows-latest]
     steps:
       - uses: actions/checkout@v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - uses: actions/cache@v2
         if: matrix.os == 'ubuntu-latest'
         with:

.github/workflows/go.yml

@@ -11,19 +11,23 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   build:
     name: Go ${{ matrix.go-version }} on ${{ matrix.os }} - healing
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        go-version: [1.17.x]
+        go-version: [1.17.x, 1.18.x]
         os: [ubuntu-latest]
     steps:
       - uses: actions/checkout@v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - uses: actions/cache@v2
         with:
           path: |

.github/workflows/iam-integrations.yaml

@@ -11,6 +11,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   iam-matrix-test:
     name: "[Go=${{ matrix.go-version }}|ldap=${{ matrix.ldap }}|etcd=${{ matrix.etcd }}|openid=${{ matrix.openid }}]"
@@ -44,6 +47,14 @@ jobs:
           - "5556:5556"
         env:
           DEX_LDAP_SERVER: "openldap:389"
+      openid2:
+        image: quay.io/minio/dex
+        ports:
+          - "5557:5557"
+        env:
+          DEX_LDAP_SERVER: "openldap:389"
+          DEX_ISSUER: "http://127.0.0.1:5557/dex"
+          DEX_WEB_HTTP: "0.0.0.0:5557"
 
     strategy:
       # When ldap, etcd or openid vars are empty below, those external servers
@@ -65,9 +76,10 @@ jobs:
 
     steps:
       - uses: actions/checkout@v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - uses: actions/cache@v2
         with:
           path: |
@@ -85,6 +97,28 @@ jobs:
           sudo sysctl net.ipv6.conf.all.disable_ipv6=0
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make test-iam
+      - name: Test with multiple OpenID providers
+        if: matrix.openid == 'http://127.0.0.1:5556/dex'
+        env:
+          LDAP_TEST_SERVER: ${{ matrix.ldap }}
+          ETCD_SERVER: ${{ matrix.etcd }}
+          OPENID_TEST_SERVER: ${{ matrix.openid }}
+          OPENID_TEST_SERVER_2: "http://127.0.0.1:5557/dex"
+        run: |
+          sudo sysctl net.ipv6.conf.all.disable_ipv6=0
+          sudo sysctl net.ipv6.conf.default.disable_ipv6=0
+          make test-iam
+      - name: Test with Access Management Plugin enabled
+        env:
+          LDAP_TEST_SERVER: ${{ matrix.ldap }}
+          ETCD_SERVER: ${{ matrix.etcd }}
+          OPENID_TEST_SERVER: ${{ matrix.openid }}
+          POLICY_PLUGIN_ENDPOINT: "http://127.0.0.1:8080"
+        run: |
+          sudo sysctl net.ipv6.conf.all.disable_ipv6=0
+          sudo sysctl net.ipv6.conf.default.disable_ipv6=0
+          go run docs/iam/access-manager-plugin.go &
+          make test-iam
       - name: Test LDAP for automatic site replication
         if: matrix.ldap == 'localhost:389'
         run: |

.github/workflows/markdown-lint.yaml

@@ -11,6 +11,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   lint:
     name: Lint all docs
@@ -22,4 +25,6 @@ jobs:
     - name: Lint all docs
       run: |
         npm install -g markdownlint-cli
-        markdownlint --fix '**/*.md' --disable MD013 MD040
+        markdownlint --fix '**/*.md' \
+          --config /home/runner/work/minio/minio/.github/markdown-lint-cfg.yaml \
+          --disable MD013 MD040

.github/workflows/replication.yaml

@@ -11,6 +11,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   replication-test:
     name: Replication Tests with Go ${{ matrix.go-version }}
@@ -22,9 +25,10 @@ jobs:
 
     steps:
       - uses: actions/checkout@v2
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
+          check-latest: true
       - uses: actions/cache@v2
         with:
           path: |

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

@@ -11,6 +11,9 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref }}
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
   build:
     name: Go ${{ matrix.go-version }} on ${{ matrix.os }}
@@ -22,10 +25,10 @@ jobs:
 
     steps:
       - uses: actions/checkout@v1
-      - uses: actions/setup-go@v2
+      - uses: actions/setup-go@v3
         with:
           go-version: ${{ matrix.go-version }}
-
+          check-latest: true
       - name: Start upgrade tests
         run: |
           make test-upgrade

.gitignore

@@ -9,8 +9,7 @@ site/
 /.idea/
 /Minio.iml
 **/access.log
-vendor/**/*.js
+vendor/
-vendor/**/*.json
 .DS_Store
 *.syso
 coverage.txt
@@ -33,3 +32,9 @@ minio.RELEASE*
 mc
 nancy
 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

@@ -25,6 +25,8 @@ linters:
     - varcheck
     - gocritic
     - gofumpt
+    - tenv
+    - durationcheck
 
 linters-settings:
   gofumpt:

Makefile

@@ -19,7 +19,7 @@ help: ## print this help
 
 getdeps: ## fetch necessary dependencies
 	@mkdir -p ${GOPATH}/bin
-	@echo "Installing golangci-lint" && curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(GOPATH)/bin v1.43.0
+	@echo "Installing golangci-lint" && curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(GOPATH)/bin v1.45.2
 	@echo "Installing msgp" && go install -v github.com/tinylib/msgp@v1.1.7-0.20211026165309-e818a1881b0e
 	@echo "Installing stringer" && go install -v golang.org/x/tools/cmd/stringer@latest
 
@@ -53,11 +53,12 @@ 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
 	@echo "Running tests for IAM (external IDP, etcd backends) with -race"
-	@CGO_ENABLED=1 go test -race -tags kqueue -v -run TestIAM* ./cmd
+	@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"
 	@(env bash $(PWD)/docs/bucket/replication/setup_3site_replication.sh)
+	@(env bash $(PWD)/docs/bucket/replication/setup_2site_existing_replication.sh)
 
 test-site-replication-ldap: install ## verify automatic site replication
 	@echo "Running tests for automatic site replication of IAM (with LDAP)"
@@ -73,15 +74,20 @@ test-site-replication-minio: install ## verify automatic site replication
 
 verify: ## verify minio various setups
 	@echo "Verifying build with race"
-	@CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
+	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
 	@(env bash $(PWD)/buildscripts/verify-build.sh)
 
 verify-healing: ## verify healing and replacing disks with minio binary
 	@echo "Verify healing build with race"
-	@CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
+	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
 	@(env bash $(PWD)/buildscripts/verify-healing.sh)
 	@(env bash $(PWD)/buildscripts/unaligned-healing.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
+	@(env bash $(PWD)/buildscripts/resolve-right-versions.sh)
+
 build: checks ## builds minio to $(PWD)
 	@echo "Building minio binary to './minio'"
 	@CGO_ENABLED=0 go build -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null

README.fips.md

@@ -0,0 +1,7 @@
+# MinIO FIPS Builds
+
+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.
+
+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

@@ -196,12 +196,6 @@ iptables -A INPUT -p tcp --dport 9000:9010 -j ACCEPT
 service iptables restart
 ```
 
-## Pre-existing data
-
-When deployed on a single drive, MinIO server lets clients access any pre-existing data in the data directory. For example, if MinIO is started with the command  `minio server /mnt/data`, any pre-existing data in the `/mnt/data` directory would be accessible to the clients.
-
-The above statement is also valid for all gateway backends.
-
 ## Test MinIO Connectivity
 
 ### Test using MinIO Console
@@ -242,16 +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 RPM/DEB installations, upgrade packages **parallelly** on all servers. Once upgraded, perform `systemctl restart minio` across all nodes in **parallel**. RPM/DEB based installations are usually automated using [`ansible`](https://github.com/minio/ansible-minio).
+- 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.
+- 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.
-- Make sure MinIO process has write access to `/opt/bin` if you plan to use `mc admin update`. This is needed for MinIO to download the latest binary from <https://dl.min.io> and save it locally for upgrades.
+- 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 in kubernetes/container environments, container environments provide their own mechanisms for container updates.
+- `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/race.sh

@@ -2,6 +2,7 @@
 
 set -e
 
+export GORACE="history_size=7"
 ## 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

@@ -0,0 +1,72 @@
+#!/bin/bash -e
+
+set -E
+set -o pipefail
+set -x
+
+WORK_DIR="$PWD/.verify-$RANDOM"
+MINIO_CONFIG_DIR="$WORK_DIR/.minio"
+MINIO=( "$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 start_minio_5drive() {
+    start_port=$1
+
+    export MINIO_ROOT_USER=minio
+    export MINIO_ROOT_PASSWORD=minio123
+    export MC_HOST_minio="http://minio:minio123@127.0.0.1:${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}"
+
+    "${WORK_DIR}/mc" cp --quiet -r "buildscripts/cicd-corpus/" "${WORK_DIR}/cicd-corpus/"
+
+    "${MINIO[@]}" --address ":$start_port" "${WORK_DIR}/cicd-corpus/disk{1...5}" > "${WORK_DIR}/server1.log" 2>&1 &
+    pid=$!
+    disown $pid
+    sleep 30
+
+    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" stat minio/bucket/testobj
+
+    pkill minio
+    sleep 3
+}
+
+function main() {
+    start_port=$(shuf -i 10000-65000 -n 1)
+
+    start_minio_5drive ${start_port}
+}
+
+function purge()
+{
+    rm -rf "$1"
+}
+
+( main "$@" )
+rv=$?
+purge "$WORK_DIR"
+exit "$rv"

cmd/admin-bucket-handlers.go

@@ -120,7 +120,7 @@ func (a adminAPIHandlers) GetBucketQuotaConfigHandler(w http.ResponseWriter, r *
 		return
 	}
 
-	config, err := globalBucketMetadataSys.GetQuotaConfig(ctx, bucket)
+	config, _, err := globalBucketMetadataSys.GetQuotaConfig(ctx, bucket)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -145,7 +145,7 @@ func (a adminAPIHandlers) SetRemoteTargetHandler(w http.ResponseWriter, r *http.
 	bucket := pathClean(vars["bucket"])
 	update := r.Form.Get("update") == "true"
 
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -274,7 +274,8 @@ func (a adminAPIHandlers) ListRemoteTargetsHandler(w http.ResponseWriter, r *htt
 	vars := mux.Vars(r)
 	bucket := pathClean(vars["bucket"])
 	arnType := vars["type"]
-	if !globalIsErasure {
+
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -314,7 +315,7 @@ func (a adminAPIHandlers) RemoveRemoteTargetHandler(w http.ResponseWriter, r *ht
 	bucket := pathClean(vars["bucket"])
 	arn := vars["arn"]
 
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}

cmd/admin-handler-utils.go

@@ -180,7 +180,7 @@ func toAdminAPIErr(ctx context.Context, err error) APIError {
 			apiErr = APIError{
 				Code:           "XMinioAdminTierBackendInUse",
 				Description:    err.Error(),
-				HTTPStatusCode: http.StatusConflict,
+				HTTPStatusCode: http.StatusBadRequest,
 			}
 		case errors.Is(err, errTierBackendNotEmpty):
 			apiErr = APIError{

cmd/admin-handlers-config-kv.go

@@ -33,7 +33,8 @@ import (
 	"github.com/minio/minio/internal/config/etcd"
 	xldap "github.com/minio/minio/internal/config/identity/ldap"
 	"github.com/minio/minio/internal/config/identity/openid"
-	"github.com/minio/minio/internal/config/policy/opa"
+	idplugin "github.com/minio/minio/internal/config/identity/plugin"
+	polplugin "github.com/minio/minio/internal/config/policy/plugin"
 	"github.com/minio/minio/internal/config/storageclass"
 	"github.com/minio/minio/internal/logger"
 	iampolicy "github.com/minio/pkg/iam/policy"
@@ -97,13 +98,14 @@ func (a adminAPIHandlers) DelConfigKVHandler(w http.ResponseWriter, r *http.Requ
 }
 
 func applyDynamic(ctx context.Context, objectAPI ObjectLayer, cfg config.Config, subSys string,
-	r *http.Request, w http.ResponseWriter) {
+	r *http.Request, w http.ResponseWriter,
+) {
 	// Apply dynamic values.
 	if err := applyDynamicConfigForSubSys(GlobalContext, objectAPI, cfg, subSys); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
-	globalNotificationSys.SignalService(serviceReloadDynamic)
+	globalNotificationSys.SignalConfigReload(subSys)
 	// Tell the client that dynamic config was applied.
 	w.Header().Set(madmin.ConfigAppliedHeader, madmin.ConfigAppliedTrue)
 }
@@ -435,14 +437,16 @@ func (a adminAPIHandlers) GetConfigHandler(w http.ResponseWriter, r *http.Reques
 				off = !cache.Enabled(kv)
 			case config.StorageClassSubSys:
 				off = !storageclass.Enabled(kv)
-			case config.PolicyOPASubSys:
+			case config.PolicyPluginSubSys:
-				off = !opa.Enabled(kv)
+				off = !polplugin.Enabled(kv)
 			case config.IdentityOpenIDSubSys:
 				off = !openid.Enabled(kv)
 			case config.IdentityLDAPSubSys:
 				off = !xldap.Enabled(kv)
 			case config.IdentityTLSSubSys:
 				off = !globalSTSTLSConfig.Enabled
+			case config.IdentityPluginSubSys:
+				off = !idplugin.Enabled(kv)
 			}
 			if off {
 				s.WriteString(config.KvComment)

cmd/admin-handlers-pools.go

@@ -58,6 +58,16 @@ func (a adminAPIHandlers) StartDecommission(w http.ResponseWriter, r *http.Reque
 		return
 	}
 
+	if ep := globalEndpoints[idx].Endpoints[0]; !ep.IsLocal {
+		for nodeIdx, proxyEp := range globalProxyEndpoints {
+			if proxyEp.Endpoint.Host == ep.Host {
+				if proxyRequestByNodeIndex(ctx, w, r, nodeIdx) {
+					return
+				}
+			}
+		}
+	}
+
 	if err := pools.Decommission(r.Context(), idx); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -96,6 +106,16 @@ func (a adminAPIHandlers) CancelDecommission(w http.ResponseWriter, r *http.Requ
 		return
 	}
 
+	if ep := globalEndpoints[idx].Endpoints[0]; !ep.IsLocal {
+		for nodeIdx, proxyEp := range globalProxyEndpoints {
+			if proxyEp.Endpoint.Host == ep.Host {
+				if proxyRequestByNodeIndex(ctx, w, r, nodeIdx) {
+					return
+				}
+			}
+		}
+	}
+
 	if err := pools.DecommissionCancel(ctx, idx); err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return

cmd/admin-handlers-site-replication.go

@@ -115,10 +115,12 @@ func (a adminAPIHandlers) SRPeerBucketOps(w http.ResponseWriter, r *http.Request
 	case madmin.MakeWithVersioningBktOp:
 		_, isLockEnabled := r.Form["lockEnabled"]
 		_, isVersioningEnabled := r.Form["versioningEnabled"]
+		_, isForceCreate := r.Form["forceCreate"]
 		opts := BucketOptions{
 			Location:          r.Form.Get("location"),
 			LockEnabled:       isLockEnabled,
 			VersioningEnabled: isVersioningEnabled,
+			ForceCreate:       isForceCreate,
 		}
 		err = globalSiteReplicationSys.PeerBucketMakeWithVersioningHandler(ctx, bucket, opts)
 	case madmin.ConfigureReplBktOp:
@@ -239,6 +241,8 @@ func (a adminAPIHandlers) SRPeerReplicateBucketItem(w http.ResponseWriter, r *ht
 				return
 			}
 		}
+	case madmin.SRBucketMetaTypeVersionConfig:
+		err = globalSiteReplicationSys.PeerBucketVersioningHandler(ctx, item.Bucket, item.Versioning)
 	case madmin.SRBucketMetaTypeTags:
 		err = globalSiteReplicationSys.PeerBucketTaggingHandler(ctx, item.Bucket, item.Tags)
 	case madmin.SRBucketMetaTypeObjectLockConfig:

cmd/admin-handlers-users-race_test.go

@@ -26,6 +26,7 @@ package cmd
 import (
 	"context"
 	"fmt"
+	"runtime"
 	"sync"
 	"testing"
 	"time"
@@ -41,11 +42,15 @@ func runAllIAMConcurrencyTests(suite *TestSuiteIAM, c *check) {
 }
 
 func TestIAMInternalIDPConcurrencyServerSuite(t *testing.T) {
+	if runtime.GOOS == globalWindowsOSName {
+		t.Skip("windows is clunky")
+	}
+
 	baseTestCases := []TestSuiteCommon{
-		// Init and run test on FS backend with signature v4.
+		// Init and run test on ErasureSD backend with signature v4.
-		{serverType: "FS", signer: signerV4},
+		{serverType: "ErasureSD", signer: signerV4},
-		// Init and run test on FS backend, with tls enabled.
+		// Init and run test on ErasureSD backend, with tls enabled.
-		{serverType: "FS", signer: signerV4, secure: true},
+		{serverType: "ErasureSD", signer: signerV4, secure: true},
 		// Init and run test on Erasure backend.
 		{serverType: "Erasure", signer: signerV4},
 		// Init and run test on ErasureSet backend.
@@ -73,7 +78,7 @@ func TestIAMInternalIDPConcurrencyServerSuite(t *testing.T) {
 }
 
 func (s *TestSuiteIAM) TestDeleteUserRace(c *check) {
-	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	ctx, cancel := context.WithTimeout(context.Background(), 90*time.Second)
 	defer cancel()
 
 	bucket := getRandomBucketName()

cmd/admin-handlers-users.go

@@ -90,7 +90,7 @@ func (a adminAPIHandlers) ListBucketUsers(w http.ResponseWriter, r *http.Request
 
 	password := cred.SecretKey
 
-	allCredentials, err := globalIAMSys.ListBucketUsers(bucket)
+	allCredentials, err := globalIAMSys.ListBucketUsers(ctx, bucket)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -124,12 +124,24 @@ func (a adminAPIHandlers) ListUsers(w http.ResponseWriter, r *http.Request) {
 
 	password := cred.SecretKey
 
-	allCredentials, err := globalIAMSys.ListUsers()
+	allCredentials, err := globalIAMSys.ListUsers(ctx)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
 
+	// Add ldap users which have mapped policies if in LDAP mode
+	// FIXME(vadmeste): move this to policy info in the future
+	ldapUsers, err := globalIAMSys.ListLDAPUsers()
+	if err != nil && err != errIAMActionNotAllowed {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+	for k, v := range ldapUsers {
+		allCredentials[k] = v
+	}
+
+	// Marshal the response
 	data, err := json.Marshal(allCredentials)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -229,6 +241,15 @@ func (a adminAPIHandlers) UpdateGroupMembers(w http.ResponseWriter, r *http.Requ
 	if updReq.IsRemove {
 		err = globalIAMSys.RemoveUsersFromGroup(ctx, updReq.Group, updReq.Members)
 	} else {
+		// Check if group already exists
+		if _, gerr := globalIAMSys.GetGroupDescription(updReq.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(updReq.Group) {
+				writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+				return
+			}
+		}
 		err = globalIAMSys.AddUsersToGroup(ctx, updReq.Group, updReq.Members)
 	}
 	if err != nil {
@@ -430,6 +451,12 @@ func (a adminAPIHandlers) AddUser(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
+	// Check if accessKey has beginning and end space characters, this only applies to new users.
+	if !exists && hasSpaceBE(accessKey) {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+		return
+	}
+
 	checkDenyOnly := false
 	if accessKey == cred.AccessKey {
 		// Check that there is no explicit deny - otherwise it's allowed
@@ -521,6 +548,12 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 		return
 	}
 
+	// service account access key cannot have space characters beginning and end of the string.
+	if hasSpaceBE(createReq.AccessKey) {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+		return
+	}
+
 	var (
 		targetUser   string
 		targetGroups []string
@@ -605,7 +638,6 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 			ConditionValues: getConditionValues(r, "", cred.AccessKey, claims),
 			IsOwner:         owner,
 			Claims:          claims,
-			DenyOnly:        true,
 		}) {
 			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
 			return
@@ -645,29 +677,6 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 		return
 	}
 
-	// Call hook for cluster-replication if the service account is not for a
-	// root user.
-	if newCred.ParentUser != globalActiveCred.AccessKey {
-		err = globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
-			Type: madmin.SRIAMItemSvcAcc,
-			SvcAccChange: &madmin.SRSvcAccChange{
-				Create: &madmin.SRSvcAccCreate{
-					Parent:        newCred.ParentUser,
-					AccessKey:     newCred.AccessKey,
-					SecretKey:     newCred.SecretKey,
-					Groups:        newCred.Groups,
-					Claims:        opts.claims,
-					SessionPolicy: createReq.Policy,
-					Status:        auth.AccountOn,
-				},
-			},
-		})
-		if err != nil {
-			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-			return
-		}
-	}
-
 	createResp := madmin.AddServiceAccountResp{
 		Credentials: madmin.Credentials{
 			AccessKey: newCred.AccessKey,
@@ -688,6 +697,29 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 	}
 
 	writeSuccessResponseJSON(w, encryptedData)
+
+	// Call hook for cluster-replication if the service account is not for a
+	// root user.
+	if newCred.ParentUser != globalActiveCred.AccessKey {
+		err = globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
+			Type: madmin.SRIAMItemSvcAcc,
+			SvcAccChange: &madmin.SRSvcAccChange{
+				Create: &madmin.SRSvcAccCreate{
+					Parent:        newCred.ParentUser,
+					AccessKey:     newCred.AccessKey,
+					SecretKey:     newCred.SecretKey,
+					Groups:        newCred.Groups,
+					Claims:        opts.claims,
+					SessionPolicy: createReq.Policy,
+					Status:        auth.AccountOn,
+				},
+			},
+		})
+		if err != nil {
+			logger.LogIf(ctx, err)
+			return
+		}
+	}
 }
 
 // UpdateServiceAccount - POST /minio/admin/v3/update-service-account
@@ -982,11 +1014,9 @@ func (a adminAPIHandlers) DeleteServiceAccount(w http.ResponseWriter, r *http.Re
 		return
 	}
 
-	svcAccount, _, err := globalIAMSys.GetServiceAccount(ctx, serviceAccount)
+	// We do not care if service account is readable or not at this point,
-	if err != nil {
+	// since this is a delete call we shall allow it to be deleted if possible.
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+	svcAccount, _, _ := globalIAMSys.GetServiceAccount(ctx, serviceAccount)
-		return
-	}
 
 	adminPrivilege := globalIAMSys.IsAllowed(iampolicy.Args{
 		AccountName:     cred.AccessKey,
@@ -1001,7 +1031,7 @@ func (a adminAPIHandlers) DeleteServiceAccount(w http.ResponseWriter, r *http.Re
 		if cred.ParentUser != "" {
 			parentUser = cred.ParentUser
 		}
-		if parentUser != svcAccount.ParentUser {
+		if svcAccount.ParentUser != "" && parentUser != svcAccount.ParentUser {
 			// The service account belongs to another user but return not
 			// found error to mitigate brute force attacks. or the
 			// serviceAccount doesn't exist.
@@ -1010,23 +1040,21 @@ func (a adminAPIHandlers) DeleteServiceAccount(w http.ResponseWriter, r *http.Re
 		}
 	}
 
-	err = globalIAMSys.DeleteServiceAccount(ctx, serviceAccount, true)
+	if err := globalIAMSys.DeleteServiceAccount(ctx, serviceAccount, true); err != nil {
-	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
 
 	// Call site replication hook - non-root user accounts are replicated.
-	if svcAccount.ParentUser != globalActiveCred.AccessKey {
+	if svcAccount.ParentUser != "" && svcAccount.ParentUser != globalActiveCred.AccessKey {
-		err = globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
+		if err := globalSiteReplicationSys.IAMChangeHook(ctx, madmin.SRIAMItem{
 			Type: madmin.SRIAMItemSvcAcc,
 			SvcAccChange: &madmin.SRSvcAccChange{
 				Delete: &madmin.SRSvcAccDelete{
 					AccessKey: serviceAccount,
 				},
 			},
-		})
+		}); err != nil {
-		if err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
 		}
@@ -1077,6 +1105,19 @@ func (a adminAPIHandlers) AccountInfoHandler(w http.ResponseWriter, r *http.Requ
 			rd = true
 		}
 
+		if globalIAMSys.IsAllowed(iampolicy.Args{
+			AccountName:     cred.AccessKey,
+			Groups:          cred.Groups,
+			Action:          iampolicy.GetBucketLocationAction,
+			BucketName:      bucketName,
+			ConditionValues: getConditionValues(r, "", cred.AccessKey, claims),
+			IsOwner:         owner,
+			ObjectName:      "",
+			Claims:          claims,
+		}) {
+			rd = true
+		}
+
 		if globalIAMSys.IsAllowed(iampolicy.Args{
 			AccountName:     cred.AccessKey,
 			Groups:          cred.Groups,
@@ -1171,8 +1212,8 @@ func (a adminAPIHandlers) AccountInfoHandler(w http.ResponseWriter, r *http.Requ
 
 			lcfg, _ := globalBucketObjectLockSys.Get(bucket.Name)
 			quota, _ := globalBucketQuotaSys.Get(ctx, bucket.Name)
-			rcfg, _ := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket.Name)
+			rcfg, _, _ := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket.Name)
-			tcfg, _ := globalBucketMetadataSys.GetTaggingConfig(bucket.Name)
+			tcfg, _, _ := globalBucketMetadataSys.GetTaggingConfig(bucket.Name)
 
 			acctInfo.Buckets = append(acctInfo.Buckets, madmin.BucketAccessInfo{
 				Name:                 bucket.Name,
@@ -1373,6 +1414,12 @@ func (a adminAPIHandlers) AddCannedPolicy(w http.ResponseWriter, r *http.Request
 	vars := mux.Vars(r)
 	policyName := vars["name"]
 
+	// Policy has space characters in begin and end reject such inputs.
+	if hasSpaceBE(policyName) {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+		return
+	}
+
 	// Error out if Content-Length is missing.
 	if r.ContentLength <= 0 {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrMissingContentLength), r.URL)

cmd/admin-handlers-users_test.go

@@ -28,6 +28,7 @@ import (
 	"net/http"
 	"net/url"
 	"os"
+	"runtime"
 	"strings"
 	"testing"
 	"time"
@@ -50,6 +51,8 @@ const (
 type TestSuiteIAM struct {
 	TestSuiteCommon
 
+	ServerTypeDescription string
+
 	// Flag to turn on tests for etcd backend IAM
 	withEtcdBackend bool
 
@@ -59,7 +62,15 @@ type TestSuiteIAM struct {
 }
 
 func newTestSuiteIAM(c TestSuiteCommon, withEtcdBackend bool) *TestSuiteIAM {
-	return &TestSuiteIAM{TestSuiteCommon: c, withEtcdBackend: withEtcdBackend}
+	etcdStr := ""
+	if withEtcdBackend {
+		etcdStr = " (with etcd backend)"
+	}
+	return &TestSuiteIAM{
+		TestSuiteCommon:       c,
+		ServerTypeDescription: fmt.Sprintf("%s%s", c.serverType, etcdStr),
+		withEtcdBackend:       withEtcdBackend,
+	}
 }
 
 func (s *TestSuiteIAM) iamSetup(c *check) {
@@ -87,6 +98,29 @@ func (s *TestSuiteIAM) iamSetup(c *check) {
 	}
 }
 
+// List of all IAM test suites (i.e. test server configuration combinations)
+// common to tests.
+var iamTestSuites = func() []*TestSuiteIAM {
+	baseTestCases := []TestSuiteCommon{
+		// Init and run test on ErasureSD backend with signature v4.
+		{serverType: "ErasureSD", signer: signerV4},
+		// Init and run test on ErasureSD backend, with tls enabled.
+		{serverType: "ErasureSD", signer: signerV4, secure: true},
+		// Init and run test on Erasure backend.
+		{serverType: "Erasure", signer: signerV4},
+		// Init and run test on ErasureSet backend.
+		{serverType: "ErasureSet", signer: signerV4},
+	}
+	testCases := []*TestSuiteIAM{}
+	for _, bt := range baseTestCases {
+		testCases = append(testCases,
+			newTestSuiteIAM(bt, false),
+			newTestSuiteIAM(bt, true),
+		)
+	}
+	return testCases
+}()
+
 const (
 	EnvTestEtcdBackend = "ETCD_SERVER"
 )
@@ -156,30 +190,12 @@ func (s *TestSuiteIAM) getUserClient(c *check, accessKey, secretKey, sessionToke
 }
 
 func TestIAMInternalIDPServerSuite(t *testing.T) {
-	baseTestCases := []TestSuiteCommon{
+	if runtime.GOOS == globalWindowsOSName {
-		// Init and run test on FS backend with signature v4.
+		t.Skip("windows is clunky disable these tests")
-		{serverType: "FS", signer: signerV4},
-		// Init and run test on FS backend, with tls enabled.
-		{serverType: "FS", signer: signerV4, secure: true},
-		// Init and run test on Erasure backend.
-		{serverType: "Erasure", signer: signerV4},
-		// Init and run test on ErasureSet backend.
-		{serverType: "ErasureSet", signer: signerV4},
-	}
-	testCases := []*TestSuiteIAM{}
-	for _, bt := range baseTestCases {
-		testCases = append(testCases,
-			newTestSuiteIAM(bt, false),
-			newTestSuiteIAM(bt, true),
-		)
-	}
-	for i, testCase := range testCases {
-		etcdStr := ""
-		if testCase.withEtcdBackend {
-			etcdStr = " (with etcd backend)"
 	}
+	for i, testCase := range iamTestSuites {
 		t.Run(
-			fmt.Sprintf("Test: %d, ServerType: %s%s", i+1, testCase.serverType, etcdStr),
+			fmt.Sprintf("Test: %d, ServerType: %s", i+1, testCase.ServerTypeDescription),
 			func(t *testing.T) {
 				suite := testCase
 				c := &check{t, testCase.serverType}
@@ -226,6 +242,7 @@ func (s *TestSuiteIAM) TestUserCreate(c *check) {
 	if err != nil {
 		c.Fatalf("unable to set policy: %v", err)
 	}
+
 	client := s.getUserClient(c, accessKey, secretKey, "")
 	err = client.MakeBucket(ctx, getRandomBucketName(), minio.MakeBucketOptions{})
 	if err != nil {
@@ -890,6 +907,9 @@ func (s *TestSuiteIAM) TestServiceAccountOpsByUser(c *check) {
 
 	// 5. Check that service account can be deleted.
 	c.assertSvcAccDeletion(ctx, s, userAdmClient, accessKey, bucket)
+
+	// 6. Check that service account cannot be created for some other user.
+	c.mustNotCreateSvcAccount(ctx, globalActiveCred.AccessKey, userAdmClient)
 }
 
 func (s *TestSuiteIAM) TestServiceAccountOpsByAdmin(c *check) {
@@ -960,6 +980,166 @@ func (s *TestSuiteIAM) TestServiceAccountOpsByAdmin(c *check) {
 	c.assertSvcAccDeletion(ctx, s, s.adm, accessKey, bucket)
 }
 
+func (s *TestSuiteIAM) SetUpAccMgmtPlugin(c *check) {
+	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
+	defer cancel()
+
+	pluginEndpoint := os.Getenv("POLICY_PLUGIN_ENDPOINT")
+	if pluginEndpoint == "" {
+		c.Skip("POLICY_PLUGIN_ENDPOINT not given - skipping.")
+	}
+
+	configCmds := []string{
+		"policy_plugin",
+		"url=" + pluginEndpoint,
+	}
+
+	_, err := s.adm.SetConfigKV(ctx, strings.Join(configCmds, " "))
+	if err != nil {
+		c.Fatalf("unable to setup access management plugin for tests: %v", err)
+	}
+
+	s.RestartIAMSuite(c)
+}
+
+// TestIAM_AMPInternalIDPServerSuite - tests for access management plugin
+func TestIAM_AMPInternalIDPServerSuite(t *testing.T) {
+	for i, testCase := range iamTestSuites {
+		t.Run(
+			fmt.Sprintf("Test: %d, ServerType: %s", i+1, testCase.ServerTypeDescription),
+			func(t *testing.T) {
+				suite := testCase
+				c := &check{t, testCase.serverType}
+
+				suite.SetUpSuite(c)
+				defer suite.TearDownSuite(c)
+
+				suite.SetUpAccMgmtPlugin(c)
+
+				suite.TestAccMgmtPlugin(c)
+			},
+		)
+	}
+}
+
+// TestAccMgmtPlugin - this test assumes that the access-management-plugin is
+// the same as the example in `docs/iam/access-manager-plugin.go` -
+// specifically, it denies only `s3:Put*` operations on non-root accounts.
+func (s *TestSuiteIAM) TestAccMgmtPlugin(c *check) {
+	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
+	defer cancel()
+
+	// 0. Check that owner is able to make-bucket.
+	bucket := getRandomBucketName()
+	err := s.client.MakeBucket(ctx, bucket, minio.MakeBucketOptions{})
+	if err != nil {
+		c.Fatalf("bucket creat error: %v", err)
+	}
+
+	// 1. Create a user.
+	accessKey, secretKey := mustGenerateCredentials(c)
+	err = s.adm.SetUser(ctx, accessKey, secretKey, madmin.AccountEnabled)
+	if err != nil {
+		c.Fatalf("Unable to set user: %v", err)
+	}
+
+	// 2. Check new user appears in listing
+	usersMap, err := s.adm.ListUsers(ctx)
+	if err != nil {
+		c.Fatalf("error listing: %v", err)
+	}
+	v, ok := usersMap[accessKey]
+	if !ok {
+		c.Fatalf("user not listed: %s", accessKey)
+	}
+	c.Assert(v.Status, madmin.AccountEnabled)
+
+	// 3. Check that user is able to make a bucket.
+	client := s.getUserClient(c, accessKey, secretKey, "")
+	err = client.MakeBucket(ctx, getRandomBucketName(), minio.MakeBucketOptions{})
+	if err != nil {
+		c.Fatalf("user not create bucket: %v", err)
+	}
+
+	// 3.1 check user has access to bucket
+	c.mustListObjects(ctx, client, bucket)
+
+	// 3.2 check that user cannot upload an object.
+	_, err = client.PutObject(ctx, bucket, "objectName", bytes.NewBuffer([]byte("some content")), 12, minio.PutObjectOptions{})
+	if err == nil {
+		c.Fatalf("user was able to upload unexpectedly")
+	}
+
+	// Create an madmin client with user creds
+	userAdmClient, err := madmin.NewWithOptions(s.endpoint, &madmin.Options{
+		Creds:  cr.NewStaticV4(accessKey, secretKey, ""),
+		Secure: s.secure,
+	})
+	if err != nil {
+		c.Fatalf("Err creating user admin client: %v", err)
+	}
+	userAdmClient.SetCustomTransport(s.TestSuiteCommon.client.Transport)
+
+	// Create svc acc
+	cr := c.mustCreateSvcAccount(ctx, accessKey, userAdmClient)
+
+	// 1. Check that svc account appears in listing
+	c.assertSvcAccAppearsInListing(ctx, userAdmClient, accessKey, cr.AccessKey)
+
+	// 2. Check that svc account info can be queried
+	c.assertSvcAccInfoQueryable(ctx, userAdmClient, accessKey, cr.AccessKey, false)
+
+	// 3. Check S3 access
+	c.assertSvcAccS3Access(ctx, s, cr, bucket)
+
+	// Check that session policies do not apply - as policy enforcement is
+	// delegated to plugin.
+	{
+		svcAK, svcSK := mustGenerateCredentials(c)
+
+		// This policy does not allow listing objects.
+		policyBytes := []byte(fmt.Sprintf(`{
+ "Version": "2012-10-17",
+ "Statement": [
+  {
+   "Effect": "Allow",
+   "Action": [
+    "s3:PutObject",
+    "s3:GetObject"
+   ],
+   "Resource": [
+    "arn:aws:s3:::%s/*"
+   ]
+  }
+ ]
+}`, bucket))
+		cr, err := userAdmClient.AddServiceAccount(ctx, madmin.AddServiceAccountReq{
+			Policy:     policyBytes,
+			TargetUser: accessKey,
+			AccessKey:  svcAK,
+			SecretKey:  svcSK,
+		})
+		if err != nil {
+			c.Fatalf("Unable to create svc acc: %v", err)
+		}
+		svcClient := s.getUserClient(c, cr.AccessKey, cr.SecretKey, "")
+		// Though the attached policy does not allow listing, it will be
+		// ignored because the plugin allows it.
+		c.mustListObjects(ctx, svcClient, bucket)
+	}
+
+	// 4. Check that service account's secret key and account status can be
+	// updated.
+	c.assertSvcAccSecretKeyAndStatusUpdate(ctx, s, userAdmClient, accessKey, bucket)
+
+	// 5. Check that service account can be deleted.
+	c.assertSvcAccDeletion(ctx, s, userAdmClient, accessKey, bucket)
+
+	// 6. Check that service account **can** be created for some other user.
+	// This is possible because of the policy enforced in the plugin.
+	c.mustCreateSvcAccount(ctx, globalActiveCred.AccessKey, userAdmClient)
+}
+
 func (c *check) mustCreateIAMUser(ctx context.Context, admClnt *madmin.AdminClient) madmin.Credentials {
 	randUser := mustGetUUID()
 	randPass := mustGetUUID()
@@ -1013,7 +1193,7 @@ func (c *check) mustNotListObjects(ctx context.Context, client *minio.Client, bu
 	res := client.ListObjects(ctx, bucket, minio.ListObjectsOptions{})
 	v, ok := <-res
 	if !ok || v.Err == nil {
-		c.Fatalf("user was able to list unexpectedly!")
+		c.Fatalf("user was able to list unexpectedly! on %s", bucket)
 	}
 }
 
@@ -1021,11 +1201,20 @@ func (c *check) mustListObjects(ctx context.Context, client *minio.Client, bucke
 	res := client.ListObjects(ctx, bucket, minio.ListObjectsOptions{})
 	v, ok := <-res
 	if ok && v.Err != nil {
-		msg := fmt.Sprintf("user was unable to list: %v", v.Err)
+		c.Fatalf("user was unable to list: %v", v.Err)
-		c.Fatalf(msg)
 	}
 }
 
+func (c *check) mustNotUpload(ctx context.Context, client *minio.Client, bucket string) {
+	_, err := client.PutObject(ctx, bucket, "some-object", bytes.NewBuffer([]byte("stuff")), 5, minio.PutObjectOptions{})
+	if e, ok := err.(minio.ErrorResponse); ok {
+		if e.Code == "AccessDenied" {
+			return
+		}
+	}
+	c.Fatalf("upload did not get an AccessDenied error - got %#v instead", err)
+}
+
 func (c *check) assertSvcAccS3Access(ctx context.Context, s *TestSuiteIAM, cr madmin.Credentials, bucket string) {
 	svcClient := s.getUserClient(c, cr.AccessKey, cr.SecretKey, "")
 	c.mustListObjects(ctx, svcClient, bucket)

cmd/admin-handlers.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
 //
@@ -18,14 +18,16 @@
 package cmd
 
 import (
+	"bytes"
 	"context"
 	crand "crypto/rand"
 	"crypto/subtle"
-	"crypto/tls"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"hash/crc32"
 	"io"
+	"io/ioutil"
 	"math/rand"
 	"net/http"
 	"net/url"
@@ -56,6 +58,7 @@ import (
 
 const (
 	maxEConfigJSONSize        = 262272
+	kubernetesVersionEndpoint = "https://kubernetes.default.svc/version"
 )
 
 // Only valid query params for mgmt admin APIs.
@@ -288,6 +291,8 @@ 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"`
@@ -499,7 +504,7 @@ func (a adminAPIHandlers) TopLocksHandler(w http.ResponseWriter, r *http.Request
 }
 
 // StartProfilingResult contains the status of the starting
-// profiling action in a given server
+// profiling action in a given server - deprecated API
 type StartProfilingResult struct {
 	NodeName string `json:"nodeName"`
 	Success  bool   `json:"success"`
@@ -593,6 +598,83 @@ func (a adminAPIHandlers) StartProfilingHandler(w http.ResponseWriter, r *http.R
 	writeSuccessResponseJSON(w, startProfilingResultInBytes)
 }
 
+// ProfileHandler - POST /minio/admin/v3/profile/?profilerType={profilerType}
+// ----------
+// Enable server profiling
+func (a adminAPIHandlers) ProfileHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "Profile")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	// Validate request signature.
+	_, adminAPIErr := checkAdminRequestAuth(ctx, r, iampolicy.ProfilingAdminAction, "")
+	if adminAPIErr != ErrNone {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(adminAPIErr), r.URL)
+		return
+	}
+
+	if globalNotificationSys == nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+	profileStr := r.Form.Get("profilerType")
+	profiles := strings.Split(profileStr, ",")
+	duration := time.Minute
+	if dstr := r.Form.Get("duration"); dstr != "" {
+		var err error
+		duration, err = time.ParseDuration(dstr)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+			return
+		}
+	}
+	// read request body
+	io.CopyN(ioutil.Discard, r.Body, 1)
+
+	globalProfilerMu.Lock()
+
+	if globalProfiler == nil {
+		globalProfiler = make(map[string]minioProfiler, 10)
+	}
+
+	// Stop profiler of all types if already running
+	for k, v := range globalProfiler {
+		v.Stop()
+		delete(globalProfiler, k)
+	}
+
+	// Start profiling on remote servers.
+	for _, profiler := range profiles {
+		globalNotificationSys.StartProfiling(profiler)
+
+		// Start profiling locally as well.
+		prof, err := startProfiler(profiler)
+		if err == nil {
+			globalProfiler[profiler] = prof
+		}
+	}
+	globalProfilerMu.Unlock()
+
+	timer := time.NewTimer(duration)
+	defer timer.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			for k, v := range globalProfiler {
+				v.Stop()
+				delete(globalProfiler, k)
+			}
+			return
+		case <-timer.C:
+			if !globalNotificationSys.DownloadProfilingData(ctx, w) {
+				writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminProfilerNotEnabled), r.URL)
+				return
+			}
+			return
+		}
+	}
+}
+
 // dummyFileInfo represents a dummy representation of a profile data file
 // present only in memory, it helps to generate the zip stream.
 type dummyFileInfo struct {
@@ -613,7 +695,7 @@ func (f dummyFileInfo) Sys() interface{}   { return f.sys }
 
 // DownloadProfilingHandler - POST /minio/admin/v3/profiling/download
 // ----------
-// Download profiling information of all nodes in a zip format
+// Download profiling information of all nodes in a zip format - deprecated API
 func (a adminAPIHandlers) DownloadProfilingHandler(w http.ResponseWriter, r *http.Request) {
 	ctx := newContext(r, w, "DownloadProfiling")
 
@@ -725,8 +807,7 @@ func (a adminAPIHandlers) HealHandler(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
-	// Check if this setup has an erasure coded backend.
+	if globalIsGateway {
-	if !globalIsErasure {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrHealNotImplemented), r.URL)
 		return
 	}
@@ -918,7 +999,7 @@ func (a adminAPIHandlers) BackgroundHealStatusHandler(w http.ResponseWriter, r *
 	}
 
 	// Check if this setup has an erasure coded backend.
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrHealNotImplemented), r.URL)
 		return
 	}
@@ -935,6 +1016,50 @@ func (a adminAPIHandlers) BackgroundHealStatusHandler(w http.ResponseWriter, r *
 	}
 }
 
+// NetperfHandler - perform mesh style network throughput test
+func (a adminAPIHandlers) NetperfHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "NetperfHandler")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	objectAPI, _ := validateAdminReq(ctx, w, r, iampolicy.HealthInfoAdminAction)
+	if objectAPI == nil {
+		return
+	}
+
+	if !globalIsDistErasure {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+
+	nsLock := objectAPI.NewNSLock(minioMetaBucket, "netperf")
+	lkctx, err := nsLock.GetLock(ctx, globalOperationTimeout)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(toAPIErrorCode(ctx, err)), r.URL)
+		return
+	}
+	defer nsLock.Unlock(lkctx.Cancel)
+
+	durationStr := r.Form.Get(peerRESTDuration)
+	duration, err := time.ParseDuration(durationStr)
+	if err != nil {
+		duration = globalNetPerfMinDuration
+	}
+
+	if duration < globalNetPerfMinDuration {
+		// We need sample size of minimum 10 secs.
+		duration = globalNetPerfMinDuration
+	}
+
+	duration = duration.Round(time.Second)
+
+	results := globalNotificationSys.Netperf(ctx, duration)
+	enc := json.NewEncoder(w)
+	if err := enc.Encode(madmin.NetperfResult{NodeResults: results}); err != nil {
+		return
+	}
+}
+
 // SpeedtestHandler - Deprecated. See ObjectSpeedtestHandler
 func (a adminAPIHandlers) SpeedtestHandler(w http.ResponseWriter, r *http.Request) {
 	a.ObjectSpeedtestHandler(w, r)
@@ -954,7 +1079,7 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 		return
 	}
 
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -984,38 +1109,30 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 		duration = time.Second * 10
 	}
 
-	// ignores any errors here.
+	sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, objectAPI, concurrent, size, autotune)
-	storageInfo, _ := objectAPI.StorageInfo(ctx)
-	capacityNeeded := uint64(concurrent * size)
-	capacity := uint64(GetTotalUsableCapacityFree(storageInfo.Disks, storageInfo))
 
-	if capacity < capacityNeeded {
+	if !sufficientCapacity {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, AdminError{
 			Code:       "XMinioSpeedtestInsufficientCapacity",
-			Message: fmt.Sprintf("not enough usable space available to perform speedtest - expected %s, got %s",
+			Message:    capacityErrMsg,
-				humanize.IBytes(capacityNeeded), humanize.IBytes(capacity)),
 			StatusCode: http.StatusInsufficientStorage,
 		}), r.URL)
 		return
 	}
 
-	// Verify if we can employ autotune without running out of capacity,
+	if autotune && !canAutotune {
-	// if we do run out of capacity, make sure to turn-off autotuning
-	// in such situations.
-	newConcurrent := concurrent + (concurrent+1)/2
-	autoTunedCapacityNeeded := uint64(newConcurrent * size)
-	if autotune && capacity < autoTunedCapacityNeeded {
-		// Turn-off auto-tuning if next possible concurrency would reach beyond disk capacity.
 		autotune = false
 	}
 
-	deleteBucket := func() {
+	bucketExists, err := makeObjectPerfBucket(ctx, objectAPI)
-		objectAPI.DeleteBucket(context.Background(), pathJoin(minioMetaBucket, "speedtest"), DeleteBucketOptions{
+	if err != nil {
-			Force:      true,
+		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
-			NoRecreate: true,
+		return
-		})
+	}
+
+	if !bucketExists {
+		defer deleteObjectPerfBucket(objectAPI)
 	}
-	defer deleteBucket()
 
 	// Freeze all incoming S3 API calls before running speedtest.
 	globalNotificationSys.ServiceFreeze(ctx, true)
@@ -1027,7 +1144,7 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 	defer keepAliveTicker.Stop()
 
 	enc := json.NewEncoder(w)
-	ch := speedTest(ctx, speedTestOpts{size, concurrent, duration, autotune, storageClass})
+	ch := objectSpeedTest(ctx, speedTestOpts{size, concurrent, duration, autotune, storageClass})
 	for {
 		select {
 		case <-ctx.Done():
@@ -1050,6 +1167,50 @@ func (a adminAPIHandlers) ObjectSpeedtestHandler(w http.ResponseWriter, r *http.
 	}
 }
 
+func makeObjectPerfBucket(ctx context.Context, objectAPI ObjectLayer) (bucketExists bool, err error) {
+	err = objectAPI.MakeBucketWithLocation(ctx, globalObjectPerfBucket, BucketOptions{})
+	if err != nil {
+		if _, ok := err.(BucketExists); !ok {
+			// Only BucketExists error can be ignored.
+			return false, err
+		}
+		bucketExists = true
+	}
+	return bucketExists, nil
+}
+
+func deleteObjectPerfBucket(objectAPI ObjectLayer) {
+	objectAPI.DeleteBucket(context.Background(), globalObjectPerfBucket, DeleteBucketOptions{
+		Force:      true,
+		NoRecreate: true,
+	})
+}
+
+func validateObjPerfOptions(ctx context.Context, objectAPI ObjectLayer, concurrent int, size int, autotune bool) (sufficientCapacity bool, canAutotune bool, capacityErrMsg string) {
+	storageInfo, _ := objectAPI.StorageInfo(ctx)
+	capacityNeeded := uint64(concurrent * size)
+	capacity := uint64(GetTotalUsableCapacityFree(storageInfo.Disks, storageInfo))
+
+	if capacity < capacityNeeded {
+		return false, false, fmt.Sprintf("not enough usable space available to perform speedtest - expected %s, got %s",
+			humanize.IBytes(capacityNeeded), humanize.IBytes(capacity))
+	}
+
+	// Verify if we can employ autotune without running out of capacity,
+	// if we do run out of capacity, make sure to turn-off autotuning
+	// in such situations.
+	if autotune {
+		newConcurrent := concurrent + (concurrent+1)/2
+		autoTunedCapacityNeeded := uint64(newConcurrent * size)
+		if capacity < autoTunedCapacityNeeded {
+			// Turn-off auto-tuning if next possible concurrency would reach beyond disk capacity.
+			return true, false, ""
+		}
+	}
+
+	return true, autotune, ""
+}
+
 // NetSpeedtestHandler - reports maximum network throughput
 func (a adminAPIHandlers) NetSpeedtestHandler(w http.ResponseWriter, r *http.Request) {
 	ctx := newContext(r, w, "NetSpeedtestHandler")
@@ -1068,7 +1229,7 @@ func (a adminAPIHandlers) DriveSpeedtestHandler(w http.ResponseWriter, r *http.R
 		return
 	}
 
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -1250,9 +1411,15 @@ func (a adminAPIHandlers) TraceHandler(w http.ResponseWriter, r *http.Request) {
 
 	peers, _ := newPeerRestClients(globalEndpoints)
 
-	globalTrace.Subscribe(traceCh, ctx.Done(), func(entry interface{}) bool {
+	traceFn := func(entry interface{}) bool {
 		return mustTrace(entry, traceOpts)
-	})
+	}
+
+	err = globalTrace.Subscribe(traceCh, ctx.Done(), traceFn)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrSlowDown), r.URL)
+		return
+	}
 
 	for _, peer := range peers {
 		if peer == nil {
@@ -1324,7 +1491,11 @@ func (a adminAPIHandlers) ConsoleLogHandler(w http.ResponseWriter, r *http.Reque
 
 	peers, _ := newPeerRestClients(globalEndpoints)
 
-	globalConsoleSys.Subscribe(logCh, ctx.Done(), node, limitLines, logKind, nil)
+	err = globalConsoleSys.Subscribe(logCh, ctx.Done(), node, limitLines, logKind, nil)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrSlowDown), r.URL)
+		return
+	}
 
 	for _, peer := range peers {
 		if peer == nil {
@@ -1538,6 +1709,7 @@ func getServerInfo(ctx context.Context, r *http.Request) madmin.InfoMessage {
 
 	buckets := madmin.Buckets{}
 	objects := madmin.Objects{}
+	versions := madmin.Versions{}
 	usage := madmin.Usage{}
 
 	objectAPI := newObjectLayerFn()
@@ -1549,6 +1721,7 @@ func getServerInfo(ctx context.Context, r *http.Request) madmin.InfoMessage {
 		if err == nil {
 			buckets = madmin.Buckets{Count: dataUsageInfo.BucketsCount}
 			objects = madmin.Objects{Count: dataUsageInfo.ObjectsTotalCount}
+			versions = madmin.Versions{Count: dataUsageInfo.VersionsTotalCount}
 			usage = madmin.Usage{Size: dataUsageInfo.ObjectsTotalSize}
 		} else {
 			buckets = madmin.Buckets{Error: err.Error()}
@@ -1597,6 +1770,7 @@ func getServerInfo(ctx context.Context, r *http.Request) madmin.InfoMessage {
 		DeploymentID: globalDeploymentID,
 		Buckets:      buckets,
 		Objects:      objects,
+		Versions:     versions,
 		Usage:        usage,
 		Services:     services,
 		Backend:      backend,
@@ -1604,6 +1778,36 @@ func getServerInfo(ctx context.Context, r *http.Request) madmin.InfoMessage {
 	}
 }
 
+func getKubernetesInfo(dctx context.Context) madmin.KubernetesInfo {
+	ctx, cancel := context.WithCancel(dctx)
+	defer cancel()
+
+	ki := madmin.KubernetesInfo{}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, kubernetesVersionEndpoint, nil)
+	if err != nil {
+		ki.Error = err.Error()
+		return ki
+	}
+
+	client := &http.Client{
+		Transport: NewGatewayHTTPTransport(),
+		Timeout:   10 * time.Second,
+	}
+
+	resp, err := client.Do(req)
+	if err != nil {
+		ki.Error = err.Error()
+		return ki
+	}
+
+	decoder := json.NewDecoder(resp.Body)
+	if err := decoder.Decode(&ki); err != nil {
+		ki.Error = err.Error()
+	}
+	return ki
+}
+
 // HealthInfoHandler - GET /minio/admin/v3/healthinfo
 // ----------
 // Get server health info
@@ -1618,13 +1822,17 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	}
 
 	query := r.Form
-	healthInfo := madmin.HealthInfo{Version: madmin.HealthInfoVersion}
+	healthInfo := madmin.HealthInfo{
+		Version: madmin.HealthInfoVersion,
+		Minio: madmin.MinioHealthInfo{
+			Info: madmin.MinioInfo{
+				DeploymentID: globalDeploymentID,
+			},
+		},
+	}
 	healthInfoCh := make(chan madmin.HealthInfo)
 
 	enc := json.NewEncoder(w)
-	partialWrite := func(oinfo madmin.HealthInfo) {
-		healthInfoCh <- oinfo
-	}
 
 	setCommonHeaders(w)
 
@@ -1640,7 +1848,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 		logger.LogIf(ctx, enc.Encode(healthInfo))
 	}
 
-	deadline := 1 * time.Hour
+	deadline := 10 * time.Second // Default deadline is 10secs for health diagnostics.
+	if query.Get(string(madmin.HealthDataTypePerfNet)) != "" ||
+		query.Get(string(madmin.HealthDataTypePerfDrive)) != "" ||
+		query.Get(string(madmin.HealthDataTypePerfObj)) != "" {
+		deadline = 1 * time.Hour
+	}
 	if dstr := r.Form.Get("deadline"); dstr != "" {
 		var err error
 		deadline, err = time.ParseDuration(dstr)
@@ -1650,9 +1863,6 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 		}
 	}
 
-	deadlinedCtx, deadlineCancel := context.WithTimeout(ctx, deadline)
-	defer deadlineCancel()
-
 	nsLock := objectAPI.NewNSLock(minioMetaBucket, "health-check-in-progress")
 	lkctx, err := nsLock.GetLock(ctx, newDynamicTimeout(deadline, deadline))
 	if err != nil { // returns a locked lock
@@ -1661,6 +1871,9 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	}
 	defer nsLock.Unlock(lkctx.Cancel)
 
+	healthCtx, healthCancel := context.WithTimeout(lkctx.Context(), deadline)
+	defer healthCancel()
+
 	hostAnonymizer := createHostAnonymizer()
 	// anonAddr - Anonymizes hosts in given input string.
 	anonAddr := func(addr string) string {
@@ -1679,13 +1892,27 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 		info.SetAddr(anonAddr(info.GetAddr()))
 	}
 
+	partialWrite := func(oinfo madmin.HealthInfo) {
+		select {
+		case healthInfoCh <- oinfo:
+		case <-healthCtx.Done():
+		}
+	}
+
+	getAndWritePlatformInfo := func() {
+		if IsKubernetes() {
+			healthInfo.Sys.KubernetesInfo = getKubernetesInfo(healthCtx)
+			partialWrite(healthInfo)
+		}
+	}
+
 	getAndWriteCPUs := func() {
 		if query.Get("syscpu") == "true" {
-			localCPUInfo := madmin.GetCPUs(deadlinedCtx, globalLocalNodeName)
+			localCPUInfo := madmin.GetCPUs(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localCPUInfo)
 			healthInfo.Sys.CPUInfo = append(healthInfo.Sys.CPUInfo, localCPUInfo)
 
-			peerCPUInfo := globalNotificationSys.GetCPUs(deadlinedCtx)
+			peerCPUInfo := globalNotificationSys.GetCPUs(healthCtx)
 			for _, cpuInfo := range peerCPUInfo {
 				anonymizeAddr(&cpuInfo)
 				healthInfo.Sys.CPUInfo = append(healthInfo.Sys.CPUInfo, cpuInfo)
@@ -1697,11 +1924,11 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWritePartitions := func() {
 		if query.Get("sysdrivehw") == "true" {
-			localPartitions := madmin.GetPartitions(deadlinedCtx, globalLocalNodeName)
+			localPartitions := madmin.GetPartitions(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localPartitions)
 			healthInfo.Sys.Partitions = append(healthInfo.Sys.Partitions, localPartitions)
 
-			peerPartitions := globalNotificationSys.GetPartitions(deadlinedCtx)
+			peerPartitions := globalNotificationSys.GetPartitions(healthCtx)
 			for _, p := range peerPartitions {
 				anonymizeAddr(&p)
 				healthInfo.Sys.Partitions = append(healthInfo.Sys.Partitions, p)
@@ -1712,11 +1939,11 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteOSInfo := func() {
 		if query.Get("sysosinfo") == "true" {
-			localOSInfo := madmin.GetOSInfo(deadlinedCtx, globalLocalNodeName)
+			localOSInfo := madmin.GetOSInfo(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localOSInfo)
 			healthInfo.Sys.OSInfo = append(healthInfo.Sys.OSInfo, localOSInfo)
 
-			peerOSInfos := globalNotificationSys.GetOSInfo(deadlinedCtx)
+			peerOSInfos := globalNotificationSys.GetOSInfo(healthCtx)
 			for _, o := range peerOSInfos {
 				anonymizeAddr(&o)
 				healthInfo.Sys.OSInfo = append(healthInfo.Sys.OSInfo, o)
@@ -1727,11 +1954,11 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteMemInfo := func() {
 		if query.Get("sysmem") == "true" {
-			localMemInfo := madmin.GetMemInfo(deadlinedCtx, globalLocalNodeName)
+			localMemInfo := madmin.GetMemInfo(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localMemInfo)
 			healthInfo.Sys.MemInfo = append(healthInfo.Sys.MemInfo, localMemInfo)
 
-			peerMemInfos := globalNotificationSys.GetMemInfo(deadlinedCtx)
+			peerMemInfos := globalNotificationSys.GetMemInfo(healthCtx)
 			for _, m := range peerMemInfos {
 				anonymizeAddr(&m)
 				healthInfo.Sys.MemInfo = append(healthInfo.Sys.MemInfo, m)
@@ -1742,12 +1969,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteSysErrors := func() {
 		if query.Get(string(madmin.HealthDataTypeSysErrors)) == "true" {
-			localSysErrors := madmin.GetSysErrors(deadlinedCtx, globalLocalNodeName)
+			localSysErrors := madmin.GetSysErrors(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localSysErrors)
 			healthInfo.Sys.SysErrs = append(healthInfo.Sys.SysErrs, localSysErrors)
 			partialWrite(healthInfo)
 
-			peerSysErrs := globalNotificationSys.GetSysErrors(deadlinedCtx)
+			peerSysErrs := globalNotificationSys.GetSysErrors(healthCtx)
 			for _, se := range peerSysErrs {
 				anonymizeAddr(&se)
 				healthInfo.Sys.SysErrs = append(healthInfo.Sys.SysErrs, se)
@@ -1758,12 +1985,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteSysConfig := func() {
 		if query.Get(string(madmin.HealthDataTypeSysConfig)) == "true" {
-			localSysConfig := madmin.GetSysConfig(deadlinedCtx, globalLocalNodeName)
+			localSysConfig := madmin.GetSysConfig(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localSysConfig)
 			healthInfo.Sys.SysConfig = append(healthInfo.Sys.SysConfig, localSysConfig)
 			partialWrite(healthInfo)
 
-			peerSysConfig := globalNotificationSys.GetSysConfig(deadlinedCtx)
+			peerSysConfig := globalNotificationSys.GetSysConfig(healthCtx)
 			for _, sc := range peerSysConfig {
 				anonymizeAddr(&sc)
 				healthInfo.Sys.SysConfig = append(healthInfo.Sys.SysConfig, sc)
@@ -1774,12 +2001,12 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteSysServices := func() {
 		if query.Get(string(madmin.HealthDataTypeSysServices)) == "true" {
-			localSysServices := madmin.GetSysServices(deadlinedCtx, globalLocalNodeName)
+			localSysServices := madmin.GetSysServices(healthCtx, globalLocalNodeName)
 			anonymizeAddr(&localSysServices)
 			healthInfo.Sys.SysServices = append(healthInfo.Sys.SysServices, localSysServices)
 			partialWrite(healthInfo)
 
-			peerSysServices := globalNotificationSys.GetSysServices(deadlinedCtx)
+			peerSysServices := globalNotificationSys.GetSysServices(healthCtx)
 			for _, ss := range peerSysServices {
 				anonymizeAddr(&ss)
 				healthInfo.Sys.SysServices = append(healthInfo.Sys.SysServices, ss)
@@ -1854,10 +2081,10 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 
 	getAndWriteProcInfo := func() {
 		if query.Get("sysprocess") == "true" {
-			localProcInfo := madmin.GetProcInfo(deadlinedCtx, globalLocalNodeName)
+			localProcInfo := madmin.GetProcInfo(healthCtx, globalLocalNodeName)
 			anonymizeProcInfo(&localProcInfo)
 			healthInfo.Sys.ProcInfo = append(healthInfo.Sys.ProcInfo, localProcInfo)
-			peerProcInfos := globalNotificationSys.GetProcInfo(deadlinedCtx)
+			peerProcInfos := globalNotificationSys.GetProcInfo(healthCtx)
 			for _, p := range peerProcInfos {
 				anonymizeProcInfo(&p)
 				healthInfo.Sys.ProcInfo = append(healthInfo.Sys.ProcInfo, p)
@@ -1883,49 +2110,98 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	}
 
 	getAndWriteDrivePerfInfo := func() {
-		if query.Get("perfdrive") == "true" {
+		if query.Get(string(madmin.HealthDataTypePerfDrive)) == "true" {
-			localDPI := getDrivePerfInfos(deadlinedCtx, globalLocalNodeName)
+			// Freeze all incoming S3 API calls before running speedtest.
-			anonymizeAddr(&localDPI)
+			globalNotificationSys.ServiceFreeze(ctx, true)
-			healthInfo.Perf.Drives = append(healthInfo.Perf.Drives, localDPI)
+			// unfreeze all incoming S3 API calls after speedtest.
-			partialWrite(healthInfo)
+			defer globalNotificationSys.ServiceFreeze(ctx, false)
 
-			perfCh := globalNotificationSys.GetDrivePerfInfos(deadlinedCtx)
+			opts := madmin.DriveSpeedTestOpts{
+				Serial:    false,
+				BlockSize: 4 * humanize.MiByte,
+				FileSize:  1 * humanize.GiByte,
+			}
+
+			localDPI := driveSpeedTest(ctx, opts)
+			healthInfo.Perf.DrivePerf = append(healthInfo.Perf.DrivePerf, localDPI)
+
+			perfCh := globalNotificationSys.DriveSpeedTest(ctx, opts)
 			for perfInfo := range perfCh {
-				anonymizeAddr(&perfInfo)
+				healthInfo.Perf.DrivePerf = append(healthInfo.Perf.DrivePerf, perfInfo)
-				healthInfo.Perf.Drives = append(healthInfo.Perf.Drives, perfInfo)
+			}
 			partialWrite(healthInfo)
 		}
 	}
+
+	getAndWriteObjPerfInfo := func() {
+		if query.Get(string(madmin.HealthDataTypePerfObj)) == "true" {
+			concurrent := 32
+			size := 64 * humanize.MiByte
+			autotune := true
+
+			sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, objectAPI, concurrent, size, autotune)
+
+			if !sufficientCapacity {
+				healthInfo.Perf.Error = capacityErrMsg
+				partialWrite(healthInfo)
+				return
 			}
 
-	anonymizeNetPerfInfo := func(npi *madmin.NetPerfInfo) {
+			if !canAutotune {
-		anonymizeAddr(npi)
+				autotune = false
-		rps := npi.RemotePeers
+			}
-		for idx, peer := range rps {
+
-			anonymizeAddr(&peer)
+			bucketExists, err := makeObjectPerfBucket(ctx, objectAPI)
-			rps[idx] = peer
+			if err != nil {
+				healthInfo.Perf.Error = "Could not make object perf bucket: " + err.Error()
+				partialWrite(healthInfo)
+				return
+			}
+
+			if !bucketExists {
+				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,
+				concurrencyStart: concurrent,
+				duration:         10 * time.Second,
+				autotune:         autotune,
+			}
+
+			perfCh := objectSpeedTest(ctx, opts)
+			for perfInfo := range perfCh {
+				healthInfo.Perf.ObjPerf = append(healthInfo.Perf.ObjPerf, perfInfo)
+			}
+			partialWrite(healthInfo)
 		}
-		npi.RemotePeers = rps
 	}
 
 	getAndWriteNetPerfInfo := func() {
-		if globalIsDistErasure && query.Get("perfnet") == "true" {
+		if query.Get(string(madmin.HealthDataTypePerfObj)) == "true" {
-			localNPI := globalNotificationSys.GetNetPerfInfo(deadlinedCtx)
+			if !globalIsDistErasure {
-			anonymizeNetPerfInfo(&localNPI)
+				return
-			healthInfo.Perf.Net = append(healthInfo.Perf.Net, localNPI)
-
-			partialWrite(healthInfo)
-
-			netInfos := globalNotificationSys.DispatchNetPerfChan(deadlinedCtx)
-			for netInfo := range netInfos {
-				anonymizeNetPerfInfo(&netInfo)
-				healthInfo.Perf.Net = append(healthInfo.Perf.Net, netInfo)
-				partialWrite(healthInfo)
 			}
 
-			ppi := globalNotificationSys.GetParallelNetPerfInfo(deadlinedCtx)
+			nsLock := objectAPI.NewNSLock(minioMetaBucket, "netperf")
-			anonymizeNetPerfInfo(&ppi)
+			lkctx, err := nsLock.GetLock(ctx, globalOperationTimeout)
-			healthInfo.Perf.NetParallel = ppi
+			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)
+				}
+			}
 			partialWrite(healthInfo)
 		}
 	}
@@ -1951,7 +2227,8 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 	go func() {
 		defer close(healthInfoCh)
 
-		partialWrite(healthInfo) // Write first message with only version populated
+		partialWrite(healthInfo) // Write first message with only version and deployment id populated
+		getAndWritePlatformInfo()
 		getAndWriteCPUs()
 		getAndWritePartitions()
 		getAndWriteOSInfo()
@@ -1959,6 +2236,7 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 		getAndWriteProcInfo()
 		getAndWriteMinioConfig()
 		getAndWriteDrivePerfInfo()
+		getAndWriteObjPerfInfo()
 		getAndWriteNetPerfInfo()
 		getAndWriteSysErrors()
 		getAndWriteSysServices()
@@ -2032,7 +2310,7 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 				return
 			}
 			w.(http.Flusher).Flush()
-		case <-deadlinedCtx.Done():
+		case <-healthCtx.Done():
 			return
 		}
 	}
@@ -2287,19 +2565,9 @@ func checkConnection(endpointStr string, timeout time.Duration) error {
 	ctx, cancel := context.WithTimeout(GlobalContext, timeout)
 	defer cancel()
 
-	client := &http.Client{Transport: &http.Transport{
+	client := &http.Client{
-		Proxy:                 http.ProxyFromEnvironment,
+		Transport: globalProxyTransport,
-		DialContext:           xhttp.NewCustomDialContext(timeout),
+	}
-		ResponseHeaderTimeout: 5 * time.Second,
-		TLSHandshakeTimeout:   5 * time.Second,
-		ExpectContinueTimeout: 5 * time.Second,
-		TLSClientConfig:       &tls.Config{RootCAs: globalRootCAs},
-		// Go net/http automatically unzip if content-type is
-		// gzip disable this feature, as we are always interested
-		// in raw stream.
-		DisableCompression: true,
-	}}
-	defer client.CloseIdleConnections()
 
 	req, err := http.NewRequestWithContext(ctx, http.MethodHead, endpointStr, nil)
 	if err != nil {
@@ -2392,8 +2660,7 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 	// of profiling data of all nodes
 	zipWriter := zip.NewWriter(encw)
 	defer zipWriter.Close()
-
+	rawDataFn := func(r io.Reader, host, disk, filename string, si StatInfo) error {
-	err = o.GetRawData(ctx, volume, file, func(r io.Reader, host, disk, filename string, si StatInfo) error {
 		// Prefix host+disk
 		filename = path.Join(host, disk, filename)
 		if si.Dir {
@@ -2404,6 +2671,10 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 			// Not, set it to default.
 			si.Mode = 0o600
 		}
+		if si.ModTime.IsZero() {
+			// Set time to now.
+			si.ModTime = time.Now()
+		}
 		header, zerr := zip.FileInfoHeader(dummyFileInfo{
 			name:    filename,
 			size:    si.Size,
@@ -2426,10 +2697,35 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 			logger.LogIf(ctx, err)
 		}
 		return nil
-	})
+	}
+	err = o.GetRawData(ctx, volume, file, rawDataFn)
+	if !errors.Is(err, errFileNotFound) {
 		logger.LogIf(ctx, err)
 	}
 
+	// save the format.json as part of inspect by default
+	if volume != minioMetaBucket && file != formatConfigFile {
+		err = o.GetRawData(ctx, minioMetaBucket, formatConfigFile, rawDataFn)
+	}
+	if !errors.Is(err, errFileNotFound) {
+		logger.LogIf(ctx, err)
+	}
+	// save args passed to inspect command
+	inspectArgs := []string{fmt.Sprintf(" Inspect path: %s%s%s\n", volume, slashSeparator, file)}
+	cmdLine := []string{"Server command line args: "}
+	for _, pool := range globalEndpoints {
+		cmdLine = append(cmdLine, pool.CmdLine)
+	}
+	cmdLine = append(cmdLine, "\n")
+	inspectArgs = append(inspectArgs, cmdLine...)
+	inspectArgsBytes := []byte(strings.Join(inspectArgs, " "))
+	if err = rawDataFn(bytes.NewReader(inspectArgsBytes), "", "", "inspect-input.txt", StatInfo{
+		Size: int64(len(inspectArgsBytes)),
+	}); err != nil {
+		logger.LogIf(ctx, err)
+	}
+}
+
 func createHostAnonymizerForFSMode() map[string]string {
 	hostAnonymizer := map[string]string{
 		globalLocalNodeName: "server1",
@@ -2473,6 +2769,7 @@ func anonymizeHost(hostAnonymizer map[string]string, endpoint Endpoint, poolNum
 	if !found {
 		// In distributed setup, anonymized addr = 'poolNum.serverNum'
 		newHost := fmt.Sprintf("pool%d.server%d", poolNum, srvrNum)
+		schemePfx := endpoint.Scheme + "://"
 
 		// Hostname
 		mapIfNotPresent(hostAnonymizer, endpoint.Hostname(), newHost)
@@ -2482,6 +2779,7 @@ func anonymizeHost(hostAnonymizer map[string]string, endpoint Endpoint, poolNum
 			// Host + port
 			newHostPort = newHost + ":" + endpoint.Port()
 			mapIfNotPresent(hostAnonymizer, endpoint.Host, newHostPort)
+			mapIfNotPresent(hostAnonymizer, schemePfx+endpoint.Host, newHostPort)
 		}
 
 		newHostPortPath := newHostPort
@@ -2490,10 +2788,11 @@ func anonymizeHost(hostAnonymizer map[string]string, endpoint Endpoint, poolNum
 			currentHostPortPath := endpoint.Host + endpoint.Path
 			newHostPortPath = newHostPort + endpoint.Path
 			mapIfNotPresent(hostAnonymizer, currentHostPortPath, newHostPortPath)
+			mapIfNotPresent(hostAnonymizer, schemePfx+currentHostPortPath, newHostPortPath)
 		}
 
 		// Full url
-		hostAnonymizer[currentURL] = endpoint.Scheme + "://" + newHostPortPath
+		hostAnonymizer[currentURL] = schemePfx + newHostPortPath
 	}
 }
 

cmd/admin-heal-ops.go

@@ -194,7 +194,6 @@ func (ahs *allHealState) periodicHealSeqsClean(ctx context.Context) {
 	for {
 		select {
 		case <-periodicTimer.C:
-			periodicTimer.Reset(time.Minute * 5)
 			now := UTCNow()
 			ahs.Lock()
 			for path, h := range ahs.healSeqMap {
@@ -203,6 +202,8 @@ func (ahs *allHealState) periodicHealSeqsClean(ctx context.Context) {
 				}
 			}
 			ahs.Unlock()
+
+			periodicTimer.Reset(time.Minute * 5)
 		case <-ctx.Done():
 			// server could be restarting - need
 			// to exit immediately
@@ -581,12 +582,7 @@ func (h *healSequence) pushHealResultItem(r madmin.HealResultItem) error {
 	// heal-results in memory and the client has not consumed it
 	// for too long.
 	unconsumedTimer := time.NewTimer(healUnconsumedTimeout)
-	defer func() {
+	defer unconsumedTimer.Stop()
-		// stop the timeout timer so it is garbage collected.
-		if !unconsumedTimer.Stop() {
-			<-unconsumedTimer.C
-		}
-	}()
 
 	var itemsLen int
 	for {
@@ -700,8 +696,9 @@ func (h *healSequence) queueHealTask(source healSource, healType madmin.HealItem
 	}
 	if source.opts != nil {
 		task.opts = *source.opts
+	} else {
+		task.opts.ScanMode = madmin.HealNormalScan
 	}
-	task.opts.ScanMode = globalHealConfig.ScanMode()
 
 	h.mutex.Lock()
 	h.scannedItemsMap[healType]++

cmd/admin-router.go

@@ -84,10 +84,12 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 			adminRouter.Methods(http.MethodPost).Path(adminVersion+"/pools/cancel").HandlerFunc(gz(httpTraceAll(adminAPI.CancelDecommission))).Queries("pool", "{pool:.*}")
 		}
 
-		// Profiling operations
+		// Profiling operations - deprecated API
 		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/profiling/start").HandlerFunc(gz(httpTraceAll(adminAPI.StartProfilingHandler))).
 			Queries("profilerType", "{profilerType:.*}")
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/profiling/download").HandlerFunc(gz(httpTraceAll(adminAPI.DownloadProfilingHandler)))
+		// Profiling operations
+		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/profile").HandlerFunc(gz(httpTraceAll(adminAPI.ProfileHandler)))
 
 		// Config KV operations.
 		if enableConfigOps {
@@ -168,7 +170,6 @@ 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:.*}")
 
-		if globalIsDistErasure || globalIsErasure {
 		// GetBucketQuotaConfig
 		adminRouter.Methods(http.MethodGet).Path(adminVersion+"/get-bucket-quota").HandlerFunc(
 			gz(httpTraceHdrs(adminAPI.GetBucketQuotaConfigHandler))).Queries("bucket", "{bucket:.*}")
@@ -211,7 +212,6 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/edit").HandlerFunc(gz(httpTraceHdrs(adminAPI.SiteReplicationEdit)))
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/peer/edit").HandlerFunc(gz(httpTraceHdrs(adminAPI.SRPeerEdit)))
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/peer/remove").HandlerFunc(gz(httpTraceHdrs(adminAPI.SRPeerRemove)))
-		}
 
 		if globalIsDistErasure {
 			// Top locks
@@ -224,7 +224,7 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		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.NetSpeedtestHandler))
+		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/speedtest/net").HandlerFunc(httpTraceHdrs(adminAPI.NetperfHandler))
 
 		// HTTP Trace
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/trace").HandlerFunc(gz(http.HandlerFunc(adminAPI.TraceHandler)))

cmd/admin-server-info.go

@@ -31,7 +31,10 @@ import (
 // local endpoints from given list of endpoints
 func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Request) madmin.ServerProperties {
 	var localEndpoints Endpoints
-	addr := r.Host
+	addr := globalLocalNodeName
+	if r != nil {
+		addr = r.Host
+	}
 	if globalIsDistErasure {
 		addr = globalLocalNodeName
 	}
@@ -40,7 +43,7 @@ func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Req
 		for _, endpoint := range ep.Endpoints {
 			nodeName := endpoint.Host
 			if nodeName == "" {
-				nodeName = r.Host
+				nodeName = addr
 			}
 			if endpoint.IsLocal {
 				// Only proceed for local endpoints
@@ -50,7 +53,7 @@ func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Req
 			}
 			_, present := network[nodeName]
 			if !present {
-				if err := isServerResolvable(endpoint, 2*time.Second); err == nil {
+				if err := isServerResolvable(endpoint, 5*time.Second); err == nil {
 					network[nodeName] = string(madmin.ItemOnline)
 				} else {
 					network[nodeName] = string(madmin.ItemOffline)

cmd/api-errors.go

@@ -131,7 +131,7 @@ const (
 	ErrReplicationNeedsVersioningError
 	ErrReplicationBucketNeedsVersioningError
 	ErrReplicationDenyEditError
-	ErrReplicationNoMatchingRuleError
+	ErrReplicationNoExistingObjects
 	ErrObjectRestoreAlreadyInProgress
 	ErrNoSuchKey
 	ErrNoSuchUpload
@@ -280,6 +280,7 @@ const (
 	ErrSiteReplicationBucketConfigError
 	ErrSiteReplicationBucketMetaError
 	ErrSiteReplicationIAMError
+	ErrSiteReplicationConfigMissing
 
 	// Bucket Quota error codes
 	ErrAdminBucketQuotaExceeded
@@ -383,6 +384,7 @@ const (
 	ErrAdminProfilerNotEnabled
 	ErrInvalidDecompressedSize
 	ErrAddUserInvalidArgument
+	ErrAdminResourceInvalidArgument
 	ErrAdminAccountNotEligible
 	ErrAccountNotEligible
 	ErrAdminServiceAccountNotFound
@@ -891,15 +893,15 @@ var errorCodes = errorCodeMap{
 		Description:    "Bandwidth limit for remote target must be atleast 100MBps",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
-	ErrReplicationNoMatchingRuleError: {
+	ErrReplicationNoExistingObjects: {
-		Code:           "XMinioReplicationNoMatchingRule",
+		Code:           "XMinioReplicationNoExistingObjects",
-		Description:    "No matching replication rule found for this object prefix",
+		Description:    "No matching ExistingsObjects rule enabled",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
 	ErrReplicationDenyEditError: {
 		Code:           "XMinioReplicationDenyEdit",
 		Description:    "Cannot alter local replication config since this server is in a cluster replication setup",
-		HTTPStatusCode: http.StatusConflict,
+		HTTPStatusCode: http.StatusBadRequest,
 	},
 	ErrBucketRemoteIdenticalToSource: {
 		Code:           "XMinioAdminRemoteIdenticalToSource",
@@ -1165,7 +1167,7 @@ var errorCodes = errorCodeMap{
 	ErrObjectExistsAsDirectory: {
 		Code:           "XMinioObjectExistsAsDirectory",
 		Description:    "Object name already exists as a directory.",
-		HTTPStatusCode: http.StatusConflict,
+		HTTPStatusCode: http.StatusBadRequest,
 	},
 	ErrInvalidObjectName: {
 		Code:           "XMinioInvalidObjectName",
@@ -1339,7 +1341,11 @@ var errorCodes = errorCodeMap{
 		Description:    "Error while replicating an IAM item",
 		HTTPStatusCode: http.StatusServiceUnavailable,
 	},
-
+	ErrSiteReplicationConfigMissing: {
+		Code:           "XMinioSiteReplicationConfigMissingError",
+		Description:    "Site not found in site replication configuration",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrMaximumExpires: {
 		Code:           "AuthorizationQueryParametersError",
 		Description:    "X-Amz-Expires must be less than a week (in seconds); that is, the given X-Amz-Expires must be less than 604800 seconds",
@@ -1825,6 +1831,11 @@ var errorCodes = errorCodeMap{
 		Description:    "User is not allowed to be same as admin access key",
 		HTTPStatusCode: http.StatusForbidden,
 	},
+	ErrAdminResourceInvalidArgument: {
+		Code:           "XMinioInvalidResource",
+		Description:    "Policy, user or group names are not allowed to begin or end with space characters",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrAdminAccountNotEligible: {
 		Code:           "XMinioInvalidIAMCredentials",
 		Description:    "The administrator key is not eligible for this operation",

cmd/api-headers.go

@@ -126,6 +126,11 @@ func setObjectHeaders(w http.ResponseWriter, objInfo ObjectInfo, rs *HTTPRangeSp
 
 	// Set all other user defined metadata.
 	for k, v := range objInfo.UserDefined {
+		// Empty values for object lock and retention can be skipped.
+		if v == "" && equals(k, xhttp.AmzObjectLockMode, xhttp.AmzObjectLockRetainUntilDate) {
+			continue
+		}
+
 		if strings.HasPrefix(strings.ToLower(k), ReservedMetadataPrefixLower) {
 			// Do not need to send any internal metadata
 			// values to client.

cmd/api-response.go

@@ -728,6 +728,14 @@ func generateMultiDeleteResponse(quiet bool, deletedObjects []DeletedObject, err
 }
 
 func writeResponse(w http.ResponseWriter, statusCode int, response []byte, mType mimeType) {
+	if statusCode == 0 {
+		statusCode = 200
+	}
+	// Similar check to http.checkWriteHeaderCode
+	if statusCode < 100 || statusCode > 999 {
+		logger.Error(fmt.Sprintf("invalid WriteHeader code %v", statusCode))
+		statusCode = http.StatusInternalServerError
+	}
 	setCommonHeaders(w)
 	if mType != mimeNone {
 		w.Header().Set(xhttp.ContentType, string(mType))
@@ -791,6 +799,12 @@ func writeErrorResponse(ctx context.Context, w http.ResponseWriter, err APIError
 		err.Description = fmt.Sprintf("The authorization header is malformed; the region is wrong; expecting '%s'.", globalSite.Region)
 	}
 
+	// Similar check to http.checkWriteHeaderCode
+	if err.HTTPStatusCode < 100 || err.HTTPStatusCode > 999 {
+		logger.Error(fmt.Sprintf("invalid WriteHeader code %v from %v", err.HTTPStatusCode, err.Code))
+		err.HTTPStatusCode = http.StatusInternalServerError
+	}
+
 	// Generate error response.
 	errorResponse := getAPIErrorResponse(ctx, err, reqURL.Path,
 		w.Header().Get(xhttp.AmzRequestID), globalDeploymentID)

cmd/api-router.go

@@ -342,7 +342,7 @@ func registerAPIRouter(router *mux.Router) {
 			collectAPIStats("getbucketnotification", maxClients(gz(httpTraceAll(api.GetBucketNotificationHandler))))).Queries("notification", "")
 		// ListenNotification
 		router.Methods(http.MethodGet).HandlerFunc(
-			collectAPIStats("listennotification", maxClients(gz(httpTraceAll(api.ListenNotificationHandler))))).Queries("events", "{events:.*}")
+			collectAPIStats("listennotification", gz(httpTraceAll(api.ListenNotificationHandler)))).Queries("events", "{events:.*}")
 		// ResetBucketReplicationStatus - MinIO extension API
 		router.Methods(http.MethodGet).HandlerFunc(
 			collectAPIStats("resetbucketreplicationstatus", maxClients(gz(httpTraceAll(api.ResetBucketReplicationStatusHandler))))).Queries("replication-reset-status", "")
@@ -474,7 +474,7 @@ func registerAPIRouter(router *mux.Router) {
 
 	// ListenNotification
 	apiRouter.Methods(http.MethodGet).Path(SlashSeparator).HandlerFunc(
-		collectAPIStats("listennotification", maxClients(gz(httpTraceAll(api.ListenNotificationHandler))))).Queries("events", "{events:.*}")
+		collectAPIStats("listennotification", gz(httpTraceAll(api.ListenNotificationHandler)))).Queries("events", "{events:.*}")
 
 	// ListBuckets
 	apiRouter.Methods(http.MethodGet).Path(SlashSeparator).HandlerFunc(

cmd/auth-handler.go

@@ -197,13 +197,7 @@ func mustGetClaimsFromToken(r *http.Request) map[string]interface{} {
 	return claims
 }
 
-// Fetch claims in the security token returned by the client.
+func getClaimsFromTokenWithSecret(token, secret string) (map[string]interface{}, error) {
-func getClaimsFromToken(token string) (map[string]interface{}, error) {
-	if token == "" {
-		claims := xjwt.NewMapClaims()
-		return claims.Map(), nil
-	}
-
 	// JWT token for x-amz-security-token is signed with admin
 	// secret key, temporary credentials become invalid if
 	// server admin credentials change. This is done to ensure
@@ -212,13 +206,19 @@ func getClaimsFromToken(token string) (map[string]interface{}, error) {
 	// hijacking the policies. We need to make sure that this is
 	// based an admin credential such that token cannot be decoded
 	// on the client side and is treated like an opaque value.
-	claims, err := auth.ExtractClaims(token, globalActiveCred.SecretKey)
+	claims, err := auth.ExtractClaims(token, secret)
+	if err != nil {
+		if subtle.ConstantTimeCompare([]byte(secret), []byte(globalActiveCred.SecretKey)) == 1 {
+			return nil, errAuthentication
+		}
+		claims, err = auth.ExtractClaims(token, globalActiveCred.SecretKey)
 		if err != nil {
 			return nil, errAuthentication
 		}
+	}
 
-	// If OPA is set, return without any further checks.
+	// If AuthZPlugin is set, return without any further checks.
-	if globalPolicyOPA != nil {
+	if newGlobalAuthZPluginFn() != nil {
 		return claims.Map(), nil
 	}
 
@@ -235,31 +235,58 @@ func getClaimsFromToken(token string) (map[string]interface{}, error) {
 			logger.LogIf(GlobalContext, err, logger.Application)
 			return nil, errAuthentication
 		}
-		claims.MapClaims[iampolicy.SessionPolicyName] = string(spBytes)
+		claims.MapClaims[sessionPolicyNameExtracted] = string(spBytes)
 	}
 
 	return claims.Map(), nil
 }
 
+// Fetch claims in the security token returned by the client.
+func getClaimsFromToken(token string) (map[string]interface{}, error) {
+	return getClaimsFromTokenWithSecret(token, globalActiveCred.SecretKey)
+}
+
 // Fetch claims in the security token returned by the client and validate the token.
 func checkClaimsFromToken(r *http.Request, cred auth.Credentials) (map[string]interface{}, APIErrorCode) {
 	token := getSessionToken(r)
 	if token != "" && cred.AccessKey == "" {
+		// x-amz-security-token is not allowed for anonymous access.
 		return nil, ErrNoAccessKey
 	}
-	if cred.IsServiceAccount() && token == "" {
+
-		token = cred.SessionToken
+	if token == "" && cred.IsTemp() {
-	}
+		// Temporary credentials should always have x-amz-security-token
-	if subtle.ConstantTimeCompare([]byte(token), []byte(cred.SessionToken)) != 1 {
 		return nil, ErrInvalidToken
 	}
-	claims, err := getClaimsFromToken(token)
+
+	if token != "" && !cred.IsTemp() {
+		// x-amz-security-token should not present for static credentials.
+		return nil, ErrInvalidToken
+	}
+
+	if cred.IsTemp() && subtle.ConstantTimeCompare([]byte(token), []byte(cred.SessionToken)) != 1 {
+		// validate token for temporary credentials only.
+		return nil, ErrInvalidToken
+	}
+
+	secret := globalActiveCred.SecretKey
+	if cred.IsServiceAccount() {
+		token = cred.SessionToken
+		secret = cred.SecretKey
+	}
+
+	if token != "" {
+		claims, err := getClaimsFromTokenWithSecret(token, secret)
 		if err != nil {
 			return nil, toAPIErrorCode(r.Context(), err)
 		}
 		return claims, ErrNone
 	}
 
+	claims := xjwt.NewMapClaims()
+	return claims.Map(), ErrNone
+}
+
 // Check request auth type verifies the incoming http request
 // - validates the request signature
 // - validates the policy action if anonymous tests bucket policies if any,

cmd/auth-handler_test.go

@@ -341,7 +341,8 @@ func mustNewSignedEmptyMD5Request(method string, urlStr string, contentLength in
 }
 
 func mustNewSignedBadMD5Request(method string, urlStr string, contentLength int64,
-	body io.ReadSeeker, t *testing.T) *http.Request {
+	body io.ReadSeeker, t *testing.T,
+) *http.Request {
 	req := mustNewRequest(method, urlStr, contentLength, body, t)
 	req.Header.Set("Content-Md5", "YWFhYWFhYWFhYWFhYWFhCg==")
 	cred := globalActiveCred

cmd/background-newdisks-heal-ops.go

@@ -263,7 +263,7 @@ func initAutoHeal(ctx context.Context, objAPI ObjectLayer) {
 	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...",
+		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.
@@ -277,14 +277,6 @@ func initAutoHeal(ctx context.Context, objAPI ObjectLayer) {
 		}
 	}
 
-	if err := bgSeq.healDiskMeta(objAPI); 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)
 }
 
@@ -320,9 +312,6 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 		case <-ctx.Done():
 			return
 		case <-diskCheckTimer.C:
-			// Reset to next interval.
-			diskCheckTimer.Reset(defaultMonitorNewDiskInterval)
-
 			var erasureSetInPoolDisksToHeal []map[int][]StorageAPI
 
 			healDisks := globalBackgroundHealState.getHealLocalDiskEndpoints()
@@ -333,7 +322,7 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 				// 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 content...",
+				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))
@@ -404,13 +393,15 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 					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.Info("Healing tracker missing on '%s', disk was swapped again on %s pool",
+								logger.LogIf(ctx, fmt.Errorf("Healing tracker missing on '%s', disk was swapped again on %s pool: %w",
-									disk, humanize.Ordinal(i+1))
+									disk, humanize.Ordinal(i+1), err))
 								tracker = newHealingTracker(disk)
 							}
 
@@ -438,12 +429,14 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 								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.LogIf(ctx, tracker.delete(ctx))
 								logger.Info("\n")
+							}
+							logger.LogIf(ctx, tracker.delete(ctx))
 
 							// Only upon success pop the healed disk.
 							globalBackgroundHealState.popHealLocalDisks(disk.Endpoint())
@@ -452,6 +445,9 @@ func monitorLocalDisksAndHeal(ctx context.Context, z *erasureServerPools, bgSeq
 				}
 			}
 			wg.Wait()
+
+			// Reset for next interval.
+			diskCheckTimer.Reset(defaultMonitorNewDiskInterval)
 		}
 	}
 }

cmd/bitrot.go

@@ -19,7 +19,6 @@ package cmd
 
 import (
 	"bytes"
-	"crypto/sha256"
 	"encoding/hex"
 	"errors"
 	"fmt"
@@ -27,6 +26,7 @@ import (
 	"io"
 
 	"github.com/minio/highwayhash"
+	"github.com/minio/minio/internal/hash/sha256"
 	"golang.org/x/crypto/blake2b"
 
 	xioutil "github.com/minio/minio/internal/ioutil"

cmd/bootstrap-peer-server.go

@@ -206,7 +206,7 @@ func verifyServerSystemConfig(ctx context.Context, endpointServerPools EndpointS
 		for _, clnt := range clnts {
 			if err := clnt.Verify(ctx, srcCfg); err != nil {
 				if !isNetworkError(err) {
-					logger.Info(fmt.Errorf("%s has incorrect configuration: %w", clnt.String(), err).Error())
+					logger.LogIf(ctx, fmt.Errorf("%s has incorrect configuration: %w", clnt.String(), err))
 				}
 				offlineEndpoints = append(offlineEndpoints, clnt.String())
 				continue

cmd/bucket-encryption-handlers.go

@@ -158,7 +158,7 @@ func (api objectAPIHandlers) GetBucketEncryptionHandler(w http.ResponseWriter, r
 		return
 	}
 
-	config, err := globalBucketMetadataSys.GetSSEConfig(bucket)
+	config, _, err := globalBucketMetadataSys.GetSSEConfig(bucket)
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -206,6 +206,15 @@ func (api objectAPIHandlers) DeleteBucketEncryptionHandler(w http.ResponseWriter
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
-
+	// Call site replication hook.
+	//
+	if err = globalSiteReplicationSys.BucketMetaHook(ctx, madmin.SRBucketMeta{
+		Type:      madmin.SRBucketMetaTypeSSEConfig,
+		Bucket:    bucket,
+		SSEConfig: nil,
+	}); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
 	writeSuccessNoContent(w)
 }

cmd/bucket-encryption.go

@@ -43,7 +43,8 @@ func (sys *BucketSSEConfigSys) Get(bucket string) (*sse.BucketSSEConfig, error)
 		return nil, BucketSSEConfigNotFound{Bucket: bucket}
 	}
 
-	return globalBucketMetadataSys.GetSSEConfig(bucket)
+	sseCfg, _, err := globalBucketMetadataSys.GetSSEConfig(bucket)
+	return sseCfg, err
 }
 
 // validateBucketSSEConfig parses bucket encryption configuration and validates if it is supported by MinIO.

cmd/bucket-handlers.go

@@ -21,7 +21,6 @@ import (
 	"bytes"
 	"context"
 	"encoding/base64"
-	"encoding/json"
 	"encoding/xml"
 	"fmt"
 	"io"
@@ -33,7 +32,6 @@ import (
 	"strconv"
 	"strings"
 	"sync"
-	"time"
 
 	"github.com/google/uuid"
 	"github.com/gorilla/mux"
@@ -364,6 +362,18 @@ func (api objectAPIHandlers) ListBucketsHandler(w http.ResponseWriter, r *http.R
 			}) {
 				bucketsInfo[n] = bucketInfo
 				n++
+			} else if globalIAMSys.IsAllowed(iampolicy.Args{
+				AccountName:     cred.AccessKey,
+				Groups:          cred.Groups,
+				Action:          iampolicy.GetBucketLocationAction,
+				BucketName:      bucketInfo.Name,
+				ConditionValues: getConditionValues(r, "", cred.AccessKey, cred.Claims),
+				IsOwner:         owner,
+				ObjectName:      "",
+				Claims:          cred.Claims,
+			}) {
+				bucketsInfo[n] = bucketInfo
+				n++
 			}
 		}
 		bucketsInfo = bucketsInfo[:n]
@@ -471,9 +481,6 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		hasLockEnabled = true
 	}
 
-	versioned := globalBucketVersioningSys.Enabled(bucket)
-	suspended := globalBucketVersioningSys.Suspended(bucket)
-
 	type deleteResult struct {
 		delInfo DeletedObject
 		errInfo DeleteError
@@ -481,8 +488,8 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 
 	deleteResults := make([]deleteResult, len(deleteObjectsReq.Objects))
 
+	vc, _ := globalBucketVersioningSys.Get(bucket)
 	oss := make([]*objSweeper, len(deleteObjectsReq.Objects))
-
 	for index, object := range deleteObjectsReq.Objects {
 		if apiErrCode := checkRequestAuthType(ctx, r, policy.DeleteObjectAction, bucket, object.ObjectName); apiErrCode != ErrNone {
 			if apiErrCode == ErrSignatureDoesNotMatch || apiErrCode == ErrInvalidAccessKeyID {
@@ -514,11 +521,11 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 
 		opts := ObjectOptions{
 			VersionID:        object.VersionID,
-			Versioned:        versioned,
+			Versioned:        vc.PrefixEnabled(object.ObjectName),
-			VersionSuspended: suspended,
+			VersionSuspended: vc.Suspended(),
 		}
 
-		if replicateDeletes || object.VersionID != "" && hasLockEnabled || !globalTierConfigMgr.Empty() {
+		if replicateDeletes || hasLockEnabled || !globalTierConfigMgr.Empty() {
 			if !globalTierConfigMgr.Empty() && object.VersionID == "" && opts.VersionSuspended {
 				opts.VersionID = nullVersionID
 			}
@@ -526,7 +533,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		}
 
 		if !globalTierConfigMgr.Empty() {
-			oss[index] = newObjSweeper(bucket, object.ObjectName).WithVersion(opts.VersionID).WithVersioning(versioned, suspended)
+			oss[index] = newObjSweeper(bucket, object.ObjectName).WithVersion(opts.VersionID).WithVersioning(opts.Versioned, opts.VersionSuspended)
 			oss[index].SetTransitionState(goi.TransitionedObject)
 		}
 
@@ -547,7 +554,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 				object.ReplicateDecisionStr = dsc.String()
 			}
 		}
-		if object.VersionID != "" && hasLockEnabled {
+		if hasLockEnabled {
 			if apiErrCode := enforceRetentionBypassForDelete(ctx, r, bucket, object, goi, gerr); apiErrCode != ErrNone {
 				apiErr := errorCodes.ToAPIErr(apiErrCode)
 				deleteResults[index].errInfo = DeleteError{
@@ -581,8 +588,8 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 
 	deleteList := toNames(objectsToDelete)
 	dObjects, errs := deleteObjectsFn(ctx, bucket, deleteList, ObjectOptions{
-		Versioned:        versioned,
+		PrefixEnabledFn:  vc.PrefixEnabled,
-		VersionSuspended: suspended,
+		VersionSuspended: vc.Suspended(),
 	})
 
 	for i := range errs {
@@ -638,23 +645,13 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 			continue
 		}
 
-		if replicateDeletes {
+		if replicateDeletes && (dobj.DeleteMarkerReplicationStatus() == replication.Pending || dobj.VersionPurgeStatus() == Pending) {
-			if dobj.DeleteMarkerReplicationStatus() == replication.Pending || dobj.VersionPurgeStatus() == Pending {
 			dv := DeletedObjectReplicationInfo{
 				DeletedObject: dobj,
 				Bucket:        bucket,
 			}
 			scheduleReplicationDelete(ctx, dv, objectAPI)
 		}
-		}
-
-	}
-
-	// Notify deleted event for objects.
-	for _, dobj := range deletedObjects {
-		if dobj.ObjectName == "" {
-			continue
-		}
 
 		eventName := event.ObjectRemovedDelete
 		objInfo := ObjectInfo{
@@ -706,13 +703,27 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 	bucket := vars["bucket"]
 
 	objectLockEnabled := false
-	if vs, found := r.Header[http.CanonicalHeaderKey("x-amz-bucket-object-lock-enabled")]; found {
+	if vs := r.Header.Get(xhttp.AmzObjectLockEnabled); len(vs) > 0 {
-		v := strings.ToLower(strings.Join(vs, ""))
+		v := strings.ToLower(vs)
-		if v != "true" && v != "false" {
+		switch v {
+		case "true", "false":
+			objectLockEnabled = v == "true"
+		default:
+			writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
+			return
+		}
+	}
+
+	forceCreate := false
+	if vs := r.Header.Get(xhttp.MinIOForceCreate); len(vs) > 0 {
+		v := strings.ToLower(vs)
+		switch v {
+		case "true", "false":
+			forceCreate = v == "true"
+		default:
 			writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrInvalidRequest), r.URL)
 			return
 		}
-		objectLockEnabled = v == "true"
 	}
 
 	if s3Error := checkRequestAuthType(ctx, r, policy.CreateBucketAction, bucket, ""); s3Error != ErrNone {
@@ -737,6 +748,7 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 	opts := BucketOptions{
 		Location:    location,
 		LockEnabled: objectLockEnabled,
+		ForceCreate: forceCreate,
 	}
 
 	if globalDNSConfig != nil {
@@ -1270,6 +1282,17 @@ func (api objectAPIHandlers) DeleteBucketHandler(w http.ResponseWriter, r *http.
 				writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrMethodNotAllowed), r.URL)
 				return
 			}
+			rcfg, err := getReplicationConfig(ctx, bucket)
+			switch {
+			case err != nil:
+				if _, ok := err.(BucketReplicationConfigNotFound); !ok {
+					writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrMethodNotAllowed), r.URL)
+					return
+				}
+			case rcfg.HasActiveRules("", true):
+				writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrMethodNotAllowed), r.URL)
+				return
+			}
 		}
 	}
 
@@ -1339,7 +1362,7 @@ func (api objectAPIHandlers) PutBucketObjectLockConfigHandler(w http.ResponseWri
 		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
 		return
 	}
-	if !globalIsErasure {
+	if globalIsGateway {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
 		return
 	}
@@ -1363,7 +1386,7 @@ func (api objectAPIHandlers) PutBucketObjectLockConfigHandler(w http.ResponseWri
 	}
 
 	// Deny object locking configuration settings on existing buckets without object lock enabled.
-	if _, err = globalBucketMetadataSys.GetObjectLockConfig(bucket); err != nil {
+	if _, _, err = globalBucketMetadataSys.GetObjectLockConfig(bucket); err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1416,7 +1439,7 @@ func (api objectAPIHandlers) GetBucketObjectLockConfigHandler(w http.ResponseWri
 		return
 	}
 
-	config, err := globalBucketMetadataSys.GetObjectLockConfig(bucket)
+	config, _, err := globalBucketMetadataSys.GetObjectLockConfig(bucket)
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -1518,7 +1541,7 @@ func (api objectAPIHandlers) GetBucketTaggingHandler(w http.ResponseWriter, r *h
 		return
 	}
 
-	config, err := globalBucketMetadataSys.GetTaggingConfig(bucket)
+	config, _, err := globalBucketMetadataSys.GetTaggingConfig(bucket)
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
@@ -1571,375 +1594,3 @@ func (api objectAPIHandlers) DeleteBucketTaggingHandler(w http.ResponseWriter, r
 	// Write success response.
 	writeSuccessResponseHeadersOnly(w)
 }
-
-// PutBucketReplicationConfigHandler - PUT Bucket replication configuration.
-// ----------
-// Add a replication configuration on the specified bucket as specified in https://docs.aws.amazon.com/AmazonS3/latest/API/API_PutBucketReplication.html
-func (api objectAPIHandlers) PutBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "PutBucketReplicationConfig")
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-	if !globalIsErasure {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
-		return
-	}
-	if s3Error := checkRequestAuthType(ctx, r, policy.PutReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	if globalSiteReplicationSys.isEnabled() {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationDenyEditError), r.URL)
-		return
-	}
-	if versioned := globalBucketVersioningSys.Enabled(bucket); !versioned {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationNeedsVersioningError), r.URL)
-		return
-	}
-	replicationConfig, err := replication.ParseConfig(io.LimitReader(r.Body, r.ContentLength))
-	if err != nil {
-		apiErr := errorCodes.ToAPIErr(ErrMalformedXML)
-		apiErr.Description = err.Error()
-		writeErrorResponse(ctx, w, apiErr, r.URL)
-		return
-	}
-	sameTarget, apiErr := validateReplicationDestination(ctx, bucket, replicationConfig)
-	if apiErr != noError {
-		writeErrorResponse(ctx, w, apiErr, r.URL)
-		return
-	}
-	// Validate the received bucket replication config
-	if err = replicationConfig.Validate(bucket, sameTarget); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	configData, err := xml.Marshal(replicationConfig)
-	if err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, configData); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
-	// Write success response.
-	writeSuccessResponseHeadersOnly(w)
-}
-
-// GetBucketReplicationConfigHandler - GET Bucket replication configuration.
-// ----------
-// Gets the replication configuration for a bucket.
-func (api objectAPIHandlers) GetBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "GetBucketReplicationConfig")
-
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	// check if user has permissions to perform this operation
-	if s3Error := checkRequestAuthType(ctx, r, policy.GetReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
-	config, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
-	if err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	configData, err := xml.Marshal(config)
-	if err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
-	// Write success response.
-	writeSuccessResponseXML(w, configData)
-}
-
-// DeleteBucketReplicationConfigHandler - DELETE Bucket replication config.
-// ----------
-func (api objectAPIHandlers) DeleteBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "DeleteBucketReplicationConfig")
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	if s3Error := checkRequestAuthType(ctx, r, policy.PutReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	if globalSiteReplicationSys.isEnabled() {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationDenyEditError), r.URL)
-		return
-	}
-	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, nil); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
-	// Write success response.
-	writeSuccessResponseHeadersOnly(w)
-}
-
-// GetBucketReplicationMetricsHandler - GET Bucket replication metrics.
-// ----------
-// Gets the replication metrics for a bucket.
-func (api objectAPIHandlers) GetBucketReplicationMetricsHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "GetBucketReplicationMetrics")
-
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	// check if user has permissions to perform this operation
-	if s3Error := checkRequestAuthType(ctx, r, policy.GetReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(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() {
-		usageInfo = dataUsageInfo.BucketsUsage[bucket]
-	}
-
-	w.Header().Set(xhttp.ContentType, string(mimeJSON))
-
-	enc := json.NewEncoder(w)
-	if err = enc.Encode(getLatestReplicationStats(bucket, usageInfo)); err != nil {
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
-	}
-}
-
-// ResetBucketReplicationStartHandler - starts a replication reset for all objects in a bucket which
-// qualify for replication and re-sync the object(s) to target, provided ExistingObjectReplication is
-// enabled for the qualifying rule. This API is a MinIO only extension provided for situations where
-// remote target is entirely lost,and previously replicated objects need to be re-synced. If resync is
-// already in progress it returns an error
-func (api objectAPIHandlers) ResetBucketReplicationStartHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "ResetBucketReplicationStart")
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-	durationStr := r.URL.Query().Get("older-than")
-	arn := r.URL.Query().Get("arn")
-	resetID := r.URL.Query().Get("reset-id")
-	if resetID == "" {
-		resetID = mustGetUUID()
-	}
-	var (
-		days time.Duration
-		err  error
-	)
-	if durationStr != "" {
-		days, err = time.ParseDuration(durationStr)
-		if err != nil {
-			writeErrorResponse(ctx, w, toAPIError(ctx, InvalidArgument{
-				Bucket: bucket,
-				Err:    fmt.Errorf("invalid query parameter older-than %s for %s : %w", durationStr, bucket, err),
-			}), r.URL)
-		}
-	}
-	resetBeforeDate := UTCNow().AddDate(0, 0, -1*int(days/24))
-
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	if s3Error := checkRequestAuthType(ctx, r, policy.ResetBucketReplicationStateAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
-	config, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
-	if err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-	if !config.HasActiveRules("", true) {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationNoMatchingRuleError), r.URL)
-		return
-	}
-	tgtArns := config.FilterTargetArns(
-		replication.ObjectOpts{
-			OpType:    replication.ResyncReplicationType,
-			TargetArn: arn,
-		})
-
-	if len(tgtArns) == 0 {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
-			Bucket: bucket,
-			Err:    fmt.Errorf("Remote target ARN %s missing or ineligible for replication resync", arn),
-		}), r.URL)
-		return
-	}
-
-	if len(tgtArns) > 1 && arn == "" {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
-			Bucket: bucket,
-			Err:    fmt.Errorf("ARN should be specified for replication reset"),
-		}), r.URL)
-		return
-	}
-	var rinfo ResyncTargetsInfo
-	target := globalBucketTargetSys.GetRemoteBucketTargetByArn(ctx, bucket, tgtArns[0])
-	target.ResetBeforeDate = UTCNow().AddDate(0, 0, -1*int(days/24))
-	target.ResetID = resetID
-	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:
-			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrReplicationRemoteConnectionError, err), r.URL)
-		default:
-			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
-		}
-	}
-	if err := startReplicationResync(ctx, bucket, arn, resetID, resetBeforeDate, objectAPI); err != nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
-			Bucket: bucket,
-			Err:    err,
-		}), r.URL)
-		return
-	}
-
-	data, err := json.Marshal(rinfo)
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
-	}
-	// Write success response.
-	writeSuccessResponseJSON(w, data)
-}
-
-// ResetBucketReplicationStatusHandler - returns the status of replication reset.
-// This API is a MinIO only extension
-func (api objectAPIHandlers) ResetBucketReplicationStatusHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "ResetBucketReplicationStatus")
-	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
-
-	vars := mux.Vars(r)
-	bucket := vars["bucket"]
-	arn := r.URL.Query().Get("arn")
-	var err error
-
-	objectAPI := api.ObjectAPI()
-	if objectAPI == nil {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	if s3Error := checkRequestAuthType(ctx, r, policy.ResetBucketReplicationStateAction, bucket, ""); s3Error != ErrNone {
-		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
-		return
-	}
-
-	// Check if bucket exists.
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket); 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
-	}
-
-	globalReplicationPool.resyncState.RLock()
-	brs, ok := globalReplicationPool.resyncState.statusMap[bucket]
-	if !ok {
-		brs, err = loadBucketResyncMetadata(ctx, bucket, objectAPI)
-		if err != nil {
-			writeErrorResponse(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
-				Bucket: bucket,
-				Err:    fmt.Errorf("No replication resync status available for %s", arn),
-			}), r.URL)
-		}
-		return
-	}
-
-	var rinfo ResyncTargetsInfo
-	for tarn, st := range brs.TargetsMap {
-		if arn != "" && tarn != arn {
-			continue
-		}
-		rinfo.Targets = append(rinfo.Targets, ResyncTarget{
-			Arn:             tarn,
-			ResetID:         st.ResyncID,
-			StartTime:       st.StartTime,
-			EndTime:         st.EndTime,
-			ResyncStatus:    st.ResyncStatus.String(),
-			ReplicatedSize:  st.ReplicatedSize,
-			ReplicatedCount: st.ReplicatedCount,
-			FailedSize:      st.FailedSize,
-			FailedCount:     st.FailedCount,
-			Bucket:          st.Bucket,
-			Object:          st.Object,
-		})
-	}
-	globalReplicationPool.resyncState.RUnlock()
-	data, err := json.Marshal(rinfo)
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
-	}
-	// Write success response.
-	writeSuccessResponseJSON(w, data)
-}

cmd/bucket-handlers_test.go

@@ -36,7 +36,8 @@ func TestRemoveBucketHandler(t *testing.T) {
 }
 
 func testRemoveBucketHandler(obj ObjectLayer, instanceType, bucketName string, apiRouter http.Handler,
-	credentials auth.Credentials, t *testing.T) {
+	credentials auth.Credentials, t *testing.T,
+) {
 	_, err := obj.PutObject(GlobalContext, bucketName, "test-object", mustGetPutObjReader(t, bytes.NewReader([]byte{}), int64(0), "", "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"), ObjectOptions{})
 	// if object upload fails stop the test.
 	if err != nil {

cmd/bucket-lifecycle-handlers_test.go

@@ -34,7 +34,8 @@ func TestBucketLifecycleWrongCredentials(t *testing.T) {
 
 // Test for authentication
 func testBucketLifecycleHandlersWrongCredentials(obj ObjectLayer, instanceType, bucketName string, apiRouter http.Handler,
-	credentials auth.Credentials, t *testing.T) {
+	credentials auth.Credentials, t *testing.T,
+) {
 	// test cases with sample input and expected output.
 	testCases := []struct {
 		method     string

cmd/bucket-lifecycle.go

@@ -247,11 +247,11 @@ func (t *transitionState) addLastDayStats(tier string, ts tierStats) {
 	t.lastDayStats[tier].addStats(ts)
 }
 
-func (t *transitionState) getDailyAllTierStats() dailyAllTierStats {
+func (t *transitionState) getDailyAllTierStats() DailyAllTierStats {
 	t.lastDayMu.RLock()
 	defer t.lastDayMu.RUnlock()
 
-	res := make(dailyAllTierStats, len(t.lastDayStats))
+	res := make(DailyAllTierStats, len(t.lastDayStats))
 	for tier, st := range t.lastDayStats {
 		res[tier] = st.clone()
 	}
@@ -330,7 +330,7 @@ const (
 // 2. when a transitioned object expires (based on an ILM rule).
 func expireTransitionedObject(ctx context.Context, objectAPI ObjectLayer, oi *ObjectInfo, lcOpts lifecycle.ObjectOpts, action expireAction) error {
 	var opts ObjectOptions
-	opts.Versioned = globalBucketVersioningSys.Enabled(oi.Bucket)
+	opts.Versioned = globalBucketVersioningSys.PrefixEnabled(oi.Bucket, oi.Name)
 	opts.VersionID = lcOpts.VersionID
 	opts.Expiration = ExpirationOptions{Expire: true}
 	switch action {
@@ -415,8 +415,8 @@ func transitionObject(ctx context.Context, objectAPI ObjectLayer, oi ObjectInfo)
 			ETag:   oi.ETag,
 		},
 		VersionID:        oi.VersionID,
-		Versioned:        globalBucketVersioningSys.Enabled(oi.Bucket),
+		Versioned:        globalBucketVersioningSys.PrefixEnabled(oi.Bucket, oi.Name),
-		VersionSuspended: globalBucketVersioningSys.Suspended(oi.Bucket),
+		VersionSuspended: globalBucketVersioningSys.PrefixSuspended(oi.Bucket, oi.Name),
 		MTime:            oi.ModTime,
 	}
 	return tier, objectAPI.TransitionObject(ctx, oi.Bucket, oi.Name, opts)
@@ -575,8 +575,8 @@ func (r *RestoreObjectRequest) validate(ctx context.Context, objAPI ObjectLayer)
 
 // postRestoreOpts returns ObjectOptions with version-id from the POST restore object request for a given bucket and object.
 func postRestoreOpts(ctx context.Context, r *http.Request, bucket, object string) (opts ObjectOptions, err error) {
-	versioned := globalBucketVersioningSys.Enabled(bucket)
+	versioned := globalBucketVersioningSys.PrefixEnabled(bucket, object)
-	versionSuspended := globalBucketVersioningSys.Suspended(bucket)
+	versionSuspended := globalBucketVersioningSys.PrefixSuspended(bucket, object)
 	vid := strings.TrimSpace(r.Form.Get(xhttp.VersionID))
 	if vid != "" && vid != nullVersionID {
 		_, err := uuid.Parse(vid)
@@ -627,8 +627,8 @@ func putRestoreOpts(bucket, object string, rreq *RestoreObjectRequest, objInfo O
 			meta[xhttp.AmzServerSideEncryption] = xhttp.AmzEncryptionAES
 		}
 		return ObjectOptions{
-			Versioned:        globalBucketVersioningSys.Enabled(bucket),
+			Versioned:        globalBucketVersioningSys.PrefixEnabled(bucket, object),
-			VersionSuspended: globalBucketVersioningSys.Suspended(bucket),
+			VersionSuspended: globalBucketVersioningSys.PrefixSuspended(bucket, object),
 			UserDefined:      meta,
 		}
 	}
@@ -640,8 +640,8 @@ func putRestoreOpts(bucket, object string, rreq *RestoreObjectRequest, objInfo O
 	}
 
 	return ObjectOptions{
-		Versioned:        globalBucketVersioningSys.Enabled(bucket),
+		Versioned:        globalBucketVersioningSys.PrefixEnabled(bucket, object),
-		VersionSuspended: globalBucketVersioningSys.Suspended(bucket),
+		VersionSuspended: globalBucketVersioningSys.PrefixSuspended(bucket, object),
 		UserDefined:      meta,
 		VersionID:        objInfo.VersionID,
 		MTime:            objInfo.ModTime,

cmd/bucket-lifecycle_test.go

@@ -203,7 +203,7 @@ func TestObjectIsRemote(t *testing.T) {
 		if got := fi.IsRemote(); got != tc.remote {
 			t.Fatalf("Test %d.a: expected %v got %v", i+1, tc.remote, got)
 		}
-		oi := fi.ToObjectInfo("bucket", "object")
+		oi := fi.ToObjectInfo("bucket", "object", false)
 		if got := oi.IsRemote(); got != tc.remote {
 			t.Fatalf("Test %d.b: expected %v got %v", i+1, tc.remote, got)
 		}

cmd/bucket-listobjects-handlers.go

@@ -19,6 +19,7 @@ package cmd
 
 import (
 	"context"
+	"fmt"
 	"net/http"
 	"strconv"
 	"strings"
@@ -26,26 +27,9 @@ import (
 	"github.com/gorilla/mux"
 	"github.com/minio/minio/internal/logger"
 
-	"github.com/minio/minio/internal/sync/errgroup"
 	"github.com/minio/pkg/bucket/policy"
 )
 
-func concurrentDecryptETag(ctx context.Context, objects []ObjectInfo) {
-	g := errgroup.WithNErrs(len(objects)).WithConcurrency(500)
-	for index := range objects {
-		index := index
-		g.Go(func() error {
-			size, err := objects[index].GetActualSize()
-			if err == nil {
-				objects[index].Size = size
-			}
-			objects[index].ETag = objects[index].GetActualETag(nil)
-			return nil
-		}, index)
-	}
-	g.Wait()
-}
-
 // Validate all the ListObjects query arguments, returns an APIErrorCode
 // if one of the args do not meet the required conditions.
 // Special conditions required by MinIO server are as below
@@ -116,7 +100,11 @@ func (api objectAPIHandlers) ListObjectVersionsHandler(w http.ResponseWriter, r
 		return
 	}
 
-	concurrentDecryptETag(ctx, listObjectVersionsInfo.Objects)
+	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
+	}
 
 	response := generateListVersionsResponse(bucket, prefix, marker, versionIDMarker, delimiter, encodingType, maxkeys, listObjectVersionsInfo)
 
@@ -178,7 +166,11 @@ func (api objectAPIHandlers) ListObjectsV2MHandler(w http.ResponseWriter, r *htt
 		return
 	}
 
-	concurrentDecryptETag(ctx, listObjectsV2Info.Objects)
+	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
+	}
 
 	// The next continuation token has id@node_index format to optimize paginated listing
 	nextContinuationToken := listObjectsV2Info.NextContinuationToken
@@ -253,7 +245,11 @@ func (api objectAPIHandlers) ListObjectsV2Handler(w http.ResponseWriter, r *http
 		return
 	}
 
-	concurrentDecryptETag(ctx, listObjectsV2Info.Objects)
+	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
+	}
 
 	response := generateListObjectsV2Response(bucket, prefix, token, listObjectsV2Info.NextContinuationToken, startAfter,
 		delimiter, encodingType, fetchOwner, listObjectsV2Info.IsTruncated,
@@ -350,7 +346,11 @@ func (api objectAPIHandlers) ListObjectsV1Handler(w http.ResponseWriter, r *http
 		return
 	}
 
-	concurrentDecryptETag(ctx, listObjectsInfo.Objects)
+	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
+	}
 
 	response := generateListObjectsV1Response(bucket, prefix, marker, delimiter, encodingType, maxKeys, listObjectsInfo)
 

cmd/bucket-metadata-sys.go

@@ -23,6 +23,7 @@ import (
 	"errors"
 	"fmt"
 	"sync"
+	"time"
 
 	"github.com/minio/madmin-go"
 	"github.com/minio/minio-go/v7/pkg/tags"
@@ -111,31 +112,29 @@ func (sys *BucketMetadataSys) Update(ctx context.Context, bucket string, configF
 	switch configFile {
 	case bucketPolicyConfig:
 		meta.PolicyConfigJSON = configData
+		meta.PolicyConfigUpdatedAt = UTCNow()
 	case bucketNotificationConfig:
 		meta.NotificationConfigXML = configData
 	case bucketLifecycleConfig:
 		meta.LifecycleConfigXML = configData
 	case bucketSSEConfig:
 		meta.EncryptionConfigXML = configData
+		meta.EncryptionConfigUpdatedAt = UTCNow()
 	case bucketTaggingConfig:
 		meta.TaggingConfigXML = configData
+		meta.TaggingConfigUpdatedAt = UTCNow()
 	case bucketQuotaConfigFile:
 		meta.QuotaConfigJSON = configData
+		meta.QuotaConfigUpdatedAt = UTCNow()
 	case objectLockConfig:
-		if !globalIsErasure && !globalIsDistErasure {
-			return NotImplemented{}
-		}
 		meta.ObjectLockConfigXML = configData
+		meta.ObjectLockConfigUpdatedAt = UTCNow()
 	case bucketVersioningConfig:
-		if !globalIsErasure && !globalIsDistErasure {
-			return NotImplemented{}
-		}
 		meta.VersioningConfigXML = configData
+		meta.VersioningConfigUpdatedAt = UTCNow()
 	case bucketReplicationConfig:
-		if !globalIsErasure && !globalIsDistErasure {
-			return NotImplemented{}
-		}
 		meta.ReplicationConfigXML = configData
+		meta.ReplicationConfigUpdatedAt = UTCNow()
 	case bucketTargetsFile:
 		meta.BucketTargetsConfigJSON, meta.BucketTargetsConfigMetaJSON, err = encryptBucketMetadata(meta.Name, configData, kms.Context{
 			bucket:            meta.Name,
@@ -186,44 +185,47 @@ func (sys *BucketMetadataSys) Get(bucket string) (BucketMetadata, error) {
 
 // GetVersioningConfig returns configured versioning config
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetVersioningConfig(bucket string) (*versioning.Versioning, error) {
+func (sys *BucketMetadataSys) GetVersioningConfig(bucket string) (*versioning.Versioning, time.Time, error) {
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
-		return nil, err
+		if errors.Is(err, errConfigNotFound) {
+			return &versioning.Versioning{XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/"}, meta.Created, nil
 		}
-	return meta.versioningConfig, nil
+		return &versioning.Versioning{XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/"}, time.Time{}, err
+	}
+	return meta.versioningConfig, meta.VersioningConfigUpdatedAt, nil
 }
 
 // GetTaggingConfig returns configured tagging config
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetTaggingConfig(bucket string) (*tags.Tags, error) {
+func (sys *BucketMetadataSys) GetTaggingConfig(bucket string) (*tags.Tags, time.Time, error) {
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
 		if errors.Is(err, errConfigNotFound) {
-			return nil, BucketTaggingNotFound{Bucket: bucket}
+			return nil, time.Time{}, BucketTaggingNotFound{Bucket: bucket}
 		}
-		return nil, err
+		return nil, time.Time{}, err
 	}
 	if meta.taggingConfig == nil {
-		return nil, BucketTaggingNotFound{Bucket: bucket}
+		return nil, time.Time{}, BucketTaggingNotFound{Bucket: bucket}
 	}
-	return meta.taggingConfig, nil
+	return meta.taggingConfig, meta.TaggingConfigUpdatedAt, nil
 }
 
 // GetObjectLockConfig returns configured object lock config
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetObjectLockConfig(bucket string) (*objectlock.Config, error) {
+func (sys *BucketMetadataSys) GetObjectLockConfig(bucket string) (*objectlock.Config, time.Time, error) {
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
 		if errors.Is(err, errConfigNotFound) {
-			return nil, BucketObjectLockConfigNotFound{Bucket: bucket}
+			return nil, time.Time{}, BucketObjectLockConfigNotFound{Bucket: bucket}
 		}
-		return nil, err
+		return nil, time.Time{}, err
 	}
 	if meta.objectLockConfig == nil {
-		return nil, BucketObjectLockConfigNotFound{Bucket: bucket}
+		return nil, time.Time{}, BucketObjectLockConfigNotFound{Bucket: bucket}
 	}
-	return meta.objectLockConfig, nil
+	return meta.objectLockConfig, meta.ObjectLockConfigUpdatedAt, nil
 }
 
 // GetLifecycleConfig returns configured lifecycle config
@@ -283,69 +285,79 @@ func (sys *BucketMetadataSys) GetNotificationConfig(bucket string) (*event.Confi
 
 // GetSSEConfig returns configured SSE config
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetSSEConfig(bucket string) (*bucketsse.BucketSSEConfig, error) {
+func (sys *BucketMetadataSys) GetSSEConfig(bucket string) (*bucketsse.BucketSSEConfig, time.Time, error) {
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
 		if errors.Is(err, errConfigNotFound) {
-			return nil, BucketSSEConfigNotFound{Bucket: bucket}
+			return nil, time.Time{}, BucketSSEConfigNotFound{Bucket: bucket}
 		}
-		return nil, err
+		return nil, time.Time{}, err
 	}
 	if meta.sseConfig == nil {
-		return nil, BucketSSEConfigNotFound{Bucket: bucket}
+		return nil, time.Time{}, BucketSSEConfigNotFound{Bucket: bucket}
 	}
-	return meta.sseConfig, nil
+	return meta.sseConfig, meta.EncryptionConfigUpdatedAt, nil
+}
+
+// CreatedAt returns the time of creation of bucket
+func (sys *BucketMetadataSys) CreatedAt(bucket string) (time.Time, error) {
+	meta, err := sys.GetConfig(GlobalContext, bucket)
+	if err != nil {
+		return time.Time{}, err
+	}
+	return meta.Created.UTC(), nil
 }
 
 // GetPolicyConfig returns configured bucket policy
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetPolicyConfig(bucket string) (*policy.Policy, error) {
+func (sys *BucketMetadataSys) GetPolicyConfig(bucket string) (*policy.Policy, time.Time, error) {
 	if globalIsGateway {
 		objAPI := newObjectLayerFn()
 		if objAPI == nil {
-			return nil, errServerNotInitialized
+			return nil, time.Time{}, errServerNotInitialized
 		}
-		return objAPI.GetBucketPolicy(GlobalContext, bucket)
+		p, err := objAPI.GetBucketPolicy(GlobalContext, bucket)
+		return p, UTCNow(), err
 	}
 
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
 		if errors.Is(err, errConfigNotFound) {
-			return nil, BucketPolicyNotFound{Bucket: bucket}
+			return nil, time.Time{}, BucketPolicyNotFound{Bucket: bucket}
 		}
-		return nil, err
+		return nil, time.Time{}, err
 	}
 	if meta.policyConfig == nil {
-		return nil, BucketPolicyNotFound{Bucket: bucket}
+		return nil, time.Time{}, BucketPolicyNotFound{Bucket: bucket}
 	}
-	return meta.policyConfig, nil
+	return meta.policyConfig, meta.PolicyConfigUpdatedAt, nil
 }
 
 // GetQuotaConfig returns configured bucket quota
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetQuotaConfig(ctx context.Context, bucket string) (*madmin.BucketQuota, error) {
+func (sys *BucketMetadataSys) GetQuotaConfig(ctx context.Context, bucket string) (*madmin.BucketQuota, time.Time, error) {
 	meta, err := sys.GetConfig(ctx, bucket)
 	if err != nil {
-		return nil, err
+		return nil, time.Time{}, err
 	}
-	return meta.quotaConfig, nil
+	return meta.quotaConfig, meta.QuotaConfigUpdatedAt, nil
 }
 
 // GetReplicationConfig returns configured bucket replication config
 // The returned object may not be modified.
-func (sys *BucketMetadataSys) GetReplicationConfig(ctx context.Context, bucket string) (*replication.Config, error) {
+func (sys *BucketMetadataSys) GetReplicationConfig(ctx context.Context, bucket string) (*replication.Config, time.Time, error) {
 	meta, err := sys.GetConfig(ctx, bucket)
 	if err != nil {
 		if errors.Is(err, errConfigNotFound) {
-			return nil, BucketReplicationConfigNotFound{Bucket: bucket}
+			return nil, time.Time{}, BucketReplicationConfigNotFound{Bucket: bucket}
 		}
-		return nil, err
+		return nil, time.Time{}, err
 	}
 
 	if meta.replicationConfig == nil {
-		return nil, BucketReplicationConfigNotFound{Bucket: bucket}
+		return nil, time.Time{}, BucketReplicationConfigNotFound{Bucket: bucket}
 	}
-	return meta.replicationConfig, nil
+	return meta.replicationConfig, meta.ReplicationConfigUpdatedAt, nil
 }
 
 // GetBucketTargetsConfig returns configured bucket targets for this bucket
@@ -353,6 +365,9 @@ func (sys *BucketMetadataSys) GetReplicationConfig(ctx context.Context, bucket s
 func (sys *BucketMetadataSys) GetBucketTargetsConfig(bucket string) (*madmin.BucketTargets, error) {
 	meta, err := sys.GetConfig(GlobalContext, bucket)
 	if err != nil {
+		if errors.Is(err, errConfigNotFound) {
+			return nil, BucketRemoteTargetNotFound{Bucket: bucket}
+		}
 		return nil, err
 	}
 	if meta.bucketTargetConfig == nil {
@@ -361,20 +376,6 @@ func (sys *BucketMetadataSys) GetBucketTargetsConfig(bucket string) (*madmin.Buc
 	return meta.bucketTargetConfig, nil
 }
 
-// GetBucketTarget returns the target for the bucket and arn.
-func (sys *BucketMetadataSys) GetBucketTarget(bucket string, arn string) (madmin.BucketTarget, error) {
-	targets, err := sys.GetBucketTargetsConfig(bucket)
-	if err != nil {
-		return madmin.BucketTarget{}, err
-	}
-	for _, t := range targets.Targets {
-		if t.Arn == arn {
-			return t, nil
-		}
-	}
-	return madmin.BucketTarget{}, errConfigNotFound
-}
-
 // GetConfig returns a specific configuration from the bucket metadata.
 // The returned object may not be modified.
 func (sys *BucketMetadataSys) GetConfig(ctx context.Context, bucket string) (BucketMetadata, error) {

cmd/bucket-metadata.go

@@ -81,6 +81,13 @@ type BucketMetadata struct {
 	ReplicationConfigXML        []byte
 	BucketTargetsConfigJSON     []byte
 	BucketTargetsConfigMetaJSON []byte
+	PolicyConfigUpdatedAt       time.Time
+	ObjectLockConfigUpdatedAt   time.Time
+	EncryptionConfigUpdatedAt   time.Time
+	TaggingConfigUpdatedAt      time.Time
+	QuotaConfigUpdatedAt        time.Time
+	ReplicationConfigUpdatedAt  time.Time
+	VersioningConfigUpdatedAt   time.Time
 
 	// Unexported fields. Must be updated atomically.
 	policyConfig           *policy.Policy
@@ -98,9 +105,10 @@ 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: UTCNow(),
+		Created: now,
 		notificationConfig: &event.Config{
 			XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/",
 		},
@@ -157,8 +165,13 @@ func loadBucketMetadata(ctx context.Context, objectAPI ObjectLayer, bucket strin
 	if err := b.convertLegacyConfigs(ctx, objectAPI); err != nil {
 		return b, err
 	}
+
 	// migrate unencrypted remote targets
-	return b, b.migrateTargetConfig(ctx, objectAPI)
+	if err = b.migrateTargetConfig(ctx, objectAPI); err != nil {
+		return b, err
+	}
+	b.defaultTimestamps()
+	return b, nil
 }
 
 // parseAllConfigs will parse all configs and populate the private fields.
@@ -347,6 +360,37 @@ func (b *BucketMetadata) convertLegacyConfigs(ctx context.Context, objectAPI Obj
 	return nil
 }
 
+// default timestamps to metadata Created timestamp if unset.
+func (b *BucketMetadata) defaultTimestamps() {
+	if b.PolicyConfigUpdatedAt.IsZero() {
+		b.PolicyConfigUpdatedAt = b.Created
+	}
+
+	if b.EncryptionConfigUpdatedAt.IsZero() {
+		b.EncryptionConfigUpdatedAt = b.Created
+	}
+
+	if b.TaggingConfigUpdatedAt.IsZero() {
+		b.TaggingConfigUpdatedAt = b.Created
+	}
+
+	if b.ObjectLockConfigUpdatedAt.IsZero() {
+		b.ObjectLockConfigUpdatedAt = b.Created
+	}
+
+	if b.QuotaConfigUpdatedAt.IsZero() {
+		b.QuotaConfigUpdatedAt = b.Created
+	}
+
+	if b.ReplicationConfigUpdatedAt.IsZero() {
+		b.ReplicationConfigUpdatedAt = b.Created
+	}
+
+	if b.VersioningConfigUpdatedAt.IsZero() {
+		b.VersioningConfigUpdatedAt = b.Created
+	}
+}
+
 // Save config to supplied ObjectLayer api.
 func (b *BucketMetadata) Save(ctx context.Context, api ObjectLayer) error {
 	if err := b.parseAllConfigs(ctx, api); err != nil {

cmd/bucket-metadata_gen.go

@@ -108,6 +108,48 @@ func (z *BucketMetadata) DecodeMsg(dc *msgp.Reader) (err error) {
 				err = msgp.WrapError(err, "BucketTargetsConfigMetaJSON")
 				return
 			}
+		case "PolicyConfigUpdatedAt":
+			z.PolicyConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "PolicyConfigUpdatedAt")
+				return
+			}
+		case "ObjectLockConfigUpdatedAt":
+			z.ObjectLockConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "ObjectLockConfigUpdatedAt")
+				return
+			}
+		case "EncryptionConfigUpdatedAt":
+			z.EncryptionConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "EncryptionConfigUpdatedAt")
+				return
+			}
+		case "TaggingConfigUpdatedAt":
+			z.TaggingConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "TaggingConfigUpdatedAt")
+				return
+			}
+		case "QuotaConfigUpdatedAt":
+			z.QuotaConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "QuotaConfigUpdatedAt")
+				return
+			}
+		case "ReplicationConfigUpdatedAt":
+			z.ReplicationConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "ReplicationConfigUpdatedAt")
+				return
+			}
+		case "VersioningConfigUpdatedAt":
+			z.VersioningConfigUpdatedAt, err = dc.ReadTime()
+			if err != nil {
+				err = msgp.WrapError(err, "VersioningConfigUpdatedAt")
+				return
+			}
 		default:
 			err = dc.Skip()
 			if err != nil {
@@ -121,9 +163,9 @@ func (z *BucketMetadata) DecodeMsg(dc *msgp.Reader) (err error) {
 
 // EncodeMsg implements msgp.Encodable
 func (z *BucketMetadata) EncodeMsg(en *msgp.Writer) (err error) {
-	// map header, size 14
+	// map header, size 21
 	// write "Name"
-	err = en.Append(0x8e, 0xa4, 0x4e, 0x61, 0x6d, 0x65)
+	err = en.Append(0xde, 0x0, 0x15, 0xa4, 0x4e, 0x61, 0x6d, 0x65)
 	if err != nil {
 		return
 	}
@@ -262,15 +304,85 @@ func (z *BucketMetadata) EncodeMsg(en *msgp.Writer) (err error) {
 		err = msgp.WrapError(err, "BucketTargetsConfigMetaJSON")
 		return
 	}
+	// write "PolicyConfigUpdatedAt"
+	err = en.Append(0xb5, 0x50, 0x6f, 0x6c, 0x69, 0x63, 0x79, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.PolicyConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "PolicyConfigUpdatedAt")
+		return
+	}
+	// write "ObjectLockConfigUpdatedAt"
+	err = en.Append(0xb9, 0x4f, 0x62, 0x6a, 0x65, 0x63, 0x74, 0x4c, 0x6f, 0x63, 0x6b, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.ObjectLockConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "ObjectLockConfigUpdatedAt")
+		return
+	}
+	// write "EncryptionConfigUpdatedAt"
+	err = en.Append(0xb9, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.EncryptionConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "EncryptionConfigUpdatedAt")
+		return
+	}
+	// write "TaggingConfigUpdatedAt"
+	err = en.Append(0xb6, 0x54, 0x61, 0x67, 0x67, 0x69, 0x6e, 0x67, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.TaggingConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "TaggingConfigUpdatedAt")
+		return
+	}
+	// write "QuotaConfigUpdatedAt"
+	err = en.Append(0xb4, 0x51, 0x75, 0x6f, 0x74, 0x61, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.QuotaConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "QuotaConfigUpdatedAt")
+		return
+	}
+	// write "ReplicationConfigUpdatedAt"
+	err = en.Append(0xba, 0x52, 0x65, 0x70, 0x6c, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.ReplicationConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "ReplicationConfigUpdatedAt")
+		return
+	}
+	// write "VersioningConfigUpdatedAt"
+	err = en.Append(0xb9, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x69, 0x6e, 0x67, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteTime(z.VersioningConfigUpdatedAt)
+	if err != nil {
+		err = msgp.WrapError(err, "VersioningConfigUpdatedAt")
+		return
+	}
 	return
 }
 
 // MarshalMsg implements msgp.Marshaler
 func (z *BucketMetadata) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.Require(b, z.Msgsize())
-	// map header, size 14
+	// map header, size 21
 	// string "Name"
-	o = append(o, 0x8e, 0xa4, 0x4e, 0x61, 0x6d, 0x65)
+	o = append(o, 0xde, 0x0, 0x15, 0xa4, 0x4e, 0x61, 0x6d, 0x65)
 	o = msgp.AppendString(o, z.Name)
 	// string "Created"
 	o = append(o, 0xa7, 0x43, 0x72, 0x65, 0x61, 0x74, 0x65, 0x64)
@@ -311,6 +423,27 @@ func (z *BucketMetadata) MarshalMsg(b []byte) (o []byte, err error) {
 	// string "BucketTargetsConfigMetaJSON"
 	o = append(o, 0xbb, 0x42, 0x75, 0x63, 0x6b, 0x65, 0x74, 0x54, 0x61, 0x72, 0x67, 0x65, 0x74, 0x73, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x4d, 0x65, 0x74, 0x61, 0x4a, 0x53, 0x4f, 0x4e)
 	o = msgp.AppendBytes(o, z.BucketTargetsConfigMetaJSON)
+	// string "PolicyConfigUpdatedAt"
+	o = append(o, 0xb5, 0x50, 0x6f, 0x6c, 0x69, 0x63, 0x79, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.PolicyConfigUpdatedAt)
+	// string "ObjectLockConfigUpdatedAt"
+	o = append(o, 0xb9, 0x4f, 0x62, 0x6a, 0x65, 0x63, 0x74, 0x4c, 0x6f, 0x63, 0x6b, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.ObjectLockConfigUpdatedAt)
+	// string "EncryptionConfigUpdatedAt"
+	o = append(o, 0xb9, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.EncryptionConfigUpdatedAt)
+	// string "TaggingConfigUpdatedAt"
+	o = append(o, 0xb6, 0x54, 0x61, 0x67, 0x67, 0x69, 0x6e, 0x67, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.TaggingConfigUpdatedAt)
+	// string "QuotaConfigUpdatedAt"
+	o = append(o, 0xb4, 0x51, 0x75, 0x6f, 0x74, 0x61, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.QuotaConfigUpdatedAt)
+	// string "ReplicationConfigUpdatedAt"
+	o = append(o, 0xba, 0x52, 0x65, 0x70, 0x6c, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.ReplicationConfigUpdatedAt)
+	// string "VersioningConfigUpdatedAt"
+	o = append(o, 0xb9, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x69, 0x6e, 0x67, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x55, 0x70, 0x64, 0x61, 0x74, 0x65, 0x64, 0x41, 0x74)
+	o = msgp.AppendTime(o, z.VersioningConfigUpdatedAt)
 	return
 }
 
@@ -416,6 +549,48 @@ func (z *BucketMetadata) UnmarshalMsg(bts []byte) (o []byte, err error) {
 				err = msgp.WrapError(err, "BucketTargetsConfigMetaJSON")
 				return
 			}
+		case "PolicyConfigUpdatedAt":
+			z.PolicyConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "PolicyConfigUpdatedAt")
+				return
+			}
+		case "ObjectLockConfigUpdatedAt":
+			z.ObjectLockConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "ObjectLockConfigUpdatedAt")
+				return
+			}
+		case "EncryptionConfigUpdatedAt":
+			z.EncryptionConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "EncryptionConfigUpdatedAt")
+				return
+			}
+		case "TaggingConfigUpdatedAt":
+			z.TaggingConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "TaggingConfigUpdatedAt")
+				return
+			}
+		case "QuotaConfigUpdatedAt":
+			z.QuotaConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "QuotaConfigUpdatedAt")
+				return
+			}
+		case "ReplicationConfigUpdatedAt":
+			z.ReplicationConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "ReplicationConfigUpdatedAt")
+				return
+			}
+		case "VersioningConfigUpdatedAt":
+			z.VersioningConfigUpdatedAt, bts, err = msgp.ReadTimeBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "VersioningConfigUpdatedAt")
+				return
+			}
 		default:
 			bts, err = msgp.Skip(bts)
 			if err != nil {
@@ -430,6 +605,6 @@ func (z *BucketMetadata) UnmarshalMsg(bts []byte) (o []byte, err error) {
 
 // Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
 func (z *BucketMetadata) Msgsize() (s int) {
-	s = 1 + 5 + msgp.StringPrefixSize + len(z.Name) + 8 + msgp.TimeSize + 12 + msgp.BoolSize + 17 + msgp.BytesPrefixSize + len(z.PolicyConfigJSON) + 22 + msgp.BytesPrefixSize + len(z.NotificationConfigXML) + 19 + msgp.BytesPrefixSize + len(z.LifecycleConfigXML) + 20 + msgp.BytesPrefixSize + len(z.ObjectLockConfigXML) + 20 + msgp.BytesPrefixSize + len(z.VersioningConfigXML) + 20 + msgp.BytesPrefixSize + len(z.EncryptionConfigXML) + 17 + msgp.BytesPrefixSize + len(z.TaggingConfigXML) + 16 + msgp.BytesPrefixSize + len(z.QuotaConfigJSON) + 21 + msgp.BytesPrefixSize + len(z.ReplicationConfigXML) + 24 + msgp.BytesPrefixSize + len(z.BucketTargetsConfigJSON) + 28 + msgp.BytesPrefixSize + len(z.BucketTargetsConfigMetaJSON)
+	s = 3 + 5 + msgp.StringPrefixSize + len(z.Name) + 8 + msgp.TimeSize + 12 + msgp.BoolSize + 17 + msgp.BytesPrefixSize + len(z.PolicyConfigJSON) + 22 + msgp.BytesPrefixSize + len(z.NotificationConfigXML) + 19 + msgp.BytesPrefixSize + len(z.LifecycleConfigXML) + 20 + msgp.BytesPrefixSize + len(z.ObjectLockConfigXML) + 20 + msgp.BytesPrefixSize + len(z.VersioningConfigXML) + 20 + msgp.BytesPrefixSize + len(z.EncryptionConfigXML) + 17 + msgp.BytesPrefixSize + len(z.TaggingConfigXML) + 16 + msgp.BytesPrefixSize + len(z.QuotaConfigJSON) + 21 + msgp.BytesPrefixSize + len(z.ReplicationConfigXML) + 24 + msgp.BytesPrefixSize + len(z.BucketTargetsConfigJSON) + 28 + msgp.BytesPrefixSize + len(z.BucketTargetsConfigMetaJSON) + 22 + msgp.TimeSize + 26 + msgp.TimeSize + 26 + msgp.TimeSize + 23 + msgp.TimeSize + 21 + msgp.TimeSize + 27 + msgp.TimeSize + 26 + msgp.TimeSize
 	return
 }

cmd/bucket-object-lock.go

@@ -44,7 +44,7 @@ func (sys *BucketObjectLockSys) Get(bucketName string) (r objectlock.Retention,
 		return r, nil
 	}
 
-	config, err := globalBucketMetadataSys.GetObjectLockConfig(bucketName)
+	config, _, err := globalBucketMetadataSys.GetObjectLockConfig(bucketName)
 	if err != nil {
 		if _, ok := err.(BucketObjectLockConfigNotFound); ok {
 			return r, nil

cmd/bucket-policy-handlers_test.go

@@ -113,7 +113,8 @@ func TestCreateBucket(t *testing.T) {
 
 // testCreateBucket - Test for calling Create Bucket and ensure we get one and only one success.
 func testCreateBucket(obj ObjectLayer, instanceType, bucketName string, apiRouter http.Handler,
-	credentials auth.Credentials, t *testing.T) {
+	credentials auth.Credentials, t *testing.T,
+) {
 	bucketName1 := fmt.Sprintf("%s-1", bucketName)
 
 	const n = 100
@@ -378,7 +379,8 @@ func TestGetBucketPolicyHandler(t *testing.T) {
 
 // testGetBucketPolicyHandler - Test for end point which fetches the access policy json of the given bucket.
 func testGetBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string, apiRouter http.Handler,
-	credentials auth.Credentials, t *testing.T) {
+	credentials auth.Credentials, t *testing.T,
+) {
 	// template for constructing HTTP request body for PUT bucket policy.
 	bucketPolicyTemplate := `{"Version":"2012-10-17","Statement":[{"Action":["s3:GetBucketLocation","s3:ListBucket"],"Effect":"Allow","Principal":{"AWS":["*"]},"Resource":["arn:aws:s3:::%s"]},{"Action":["s3:GetObject"],"Effect":"Allow","Principal":{"AWS":["*"]},"Resource":["arn:aws:s3:::%s/this*"]}]}`
 

cmd/bucket-policy.go

@@ -38,7 +38,8 @@ type PolicySys struct{}
 
 // Get returns stored bucket policy
 func (sys *PolicySys) Get(bucket string) (*policy.Policy, error) {
-	return globalBucketMetadataSys.GetPolicyConfig(bucket)
+	policy, _, err := globalBucketMetadataSys.GetPolicyConfig(bucket)
+	return policy, err
 }
 
 // IsAllowed - checks given policy args is allowed to continue the Rest API.

cmd/bucket-quota.go

@@ -42,8 +42,8 @@ func (sys *BucketQuotaSys) Get(ctx context.Context, bucketName string) (*madmin.
 		}
 		return &madmin.BucketQuota{}, nil
 	}
-
+	qCfg, _, err := globalBucketMetadataSys.GetQuotaConfig(ctx, bucketName)
-	return globalBucketMetadataSys.GetQuotaConfig(ctx, bucketName)
+	return qCfg, err
 }
 
 // NewBucketQuotaSys returns initialized BucketQuotaSys

cmd/bucket-replication-handlers.go

@@ -0,0 +1,404 @@
+// 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 (
+	"encoding/json"
+	"encoding/xml"
+	"fmt"
+	"io"
+	"net/http"
+	"time"
+
+	"github.com/gorilla/mux"
+	"github.com/minio/minio/internal/bucket/replication"
+	xhttp "github.com/minio/minio/internal/http"
+	"github.com/minio/minio/internal/logger"
+	"github.com/minio/pkg/bucket/policy"
+)
+
+// PutBucketReplicationConfigHandler - PUT Bucket replication configuration.
+// ----------
+// Add a replication configuration on the specified bucket as specified in https://docs.aws.amazon.com/AmazonS3/latest/API/API_PutBucketReplication.html
+func (api objectAPIHandlers) PutBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "PutBucketReplicationConfig")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+	if globalIsGateway {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrNotImplemented), r.URL)
+		return
+	}
+	if s3Error := checkRequestAuthType(ctx, r, policy.PutReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	if versioned := globalBucketVersioningSys.Enabled(bucket); !versioned {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationNeedsVersioningError), r.URL)
+		return
+	}
+	replicationConfig, err := replication.ParseConfig(io.LimitReader(r.Body, r.ContentLength))
+	if err != nil {
+		apiErr := errorCodes.ToAPIErr(ErrMalformedXML)
+		apiErr.Description = err.Error()
+		writeErrorResponse(ctx, w, apiErr, r.URL)
+		return
+	}
+	sameTarget, apiErr := validateReplicationDestination(ctx, bucket, replicationConfig, true)
+	if apiErr != noError {
+		writeErrorResponse(ctx, w, apiErr, r.URL)
+		return
+	}
+	// Validate the received bucket replication config
+	if err = replicationConfig.Validate(bucket, sameTarget); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	configData, err := xml.Marshal(replicationConfig)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	if err = globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, configData); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	// Write success response.
+	writeSuccessResponseHeadersOnly(w)
+}
+
+// GetBucketReplicationConfigHandler - GET Bucket replication configuration.
+// ----------
+// Gets the replication configuration for a bucket.
+func (api objectAPIHandlers) GetBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "GetBucketReplicationConfig")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	// check if user has permissions to perform this operation
+	if s3Error := checkRequestAuthType(ctx, r, policy.GetReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	config, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	configData, err := xml.Marshal(config)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	// Write success response.
+	writeSuccessResponseXML(w, configData)
+}
+
+// DeleteBucketReplicationConfigHandler - DELETE Bucket replication config.
+// ----------
+func (api objectAPIHandlers) DeleteBucketReplicationConfigHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "DeleteBucketReplicationConfig")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	if s3Error := checkRequestAuthType(ctx, r, policy.PutReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	if globalSiteReplicationSys.isEnabled() {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationDenyEditError), r.URL)
+		return
+	}
+	if err := globalBucketMetadataSys.Update(ctx, bucket, bucketReplicationConfig, nil); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	// Write success response.
+	writeSuccessResponseHeadersOnly(w)
+}
+
+// GetBucketReplicationMetricsHandler - GET Bucket replication metrics.
+// ----------
+// Gets the replication metrics for a bucket.
+func (api objectAPIHandlers) GetBucketReplicationMetricsHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "GetBucketReplicationMetrics")
+
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	// check if user has permissions to perform this operation
+	if s3Error := checkRequestAuthType(ctx, r, policy.GetReplicationConfigurationAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(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() {
+		usageInfo = dataUsageInfo.BucketsUsage[bucket]
+	}
+
+	w.Header().Set(xhttp.ContentType, string(mimeJSON))
+
+	enc := json.NewEncoder(w)
+	if err = enc.Encode(getLatestReplicationStats(bucket, usageInfo)); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+}
+
+// ResetBucketReplicationStartHandler - starts a replication reset for all objects in a bucket which
+// qualify for replication and re-sync the object(s) to target, provided ExistingObjectReplication is
+// enabled for the qualifying rule. This API is a MinIO only extension provided for situations where
+// remote target is entirely lost,and previously replicated objects need to be re-synced. If resync is
+// already in progress it returns an error
+func (api objectAPIHandlers) ResetBucketReplicationStartHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ResetBucketReplicationStart")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+	durationStr := r.URL.Query().Get("older-than")
+	arn := r.URL.Query().Get("arn")
+	resetID := r.URL.Query().Get("reset-id")
+	if resetID == "" {
+		resetID = mustGetUUID()
+	}
+	var (
+		days time.Duration
+		err  error
+	)
+	if durationStr != "" {
+		days, err = time.ParseDuration(durationStr)
+		if err != nil {
+			writeErrorResponse(ctx, w, toAPIError(ctx, InvalidArgument{
+				Bucket: bucket,
+				Err:    fmt.Errorf("invalid query parameter older-than %s for %s : %w", durationStr, bucket, err),
+			}), r.URL)
+		}
+	}
+	resetBeforeDate := UTCNow().AddDate(0, 0, -1*int(days/24))
+
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	if s3Error := checkRequestAuthType(ctx, r, policy.ResetBucketReplicationStateAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	config, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	if !config.HasExistingObjectReplication(arn) {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrReplicationNoExistingObjects), r.URL)
+		return
+	}
+
+	tgtArns := config.FilterTargetArns(
+		replication.ObjectOpts{
+			OpType:    replication.ResyncReplicationType,
+			TargetArn: arn,
+		})
+
+	if len(tgtArns) == 0 {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
+			Bucket: bucket,
+			Err:    fmt.Errorf("Remote target ARN %s missing or ineligible for replication resync", arn),
+		}), r.URL)
+		return
+	}
+
+	if len(tgtArns) > 1 && arn == "" {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
+			Bucket: bucket,
+			Err:    fmt.Errorf("ARN should be specified for replication reset"),
+		}), r.URL)
+		return
+	}
+	var rinfo ResyncTargetsInfo
+	target := globalBucketTargetSys.GetRemoteBucketTargetByArn(ctx, bucket, tgtArns[0])
+	target.ResetBeforeDate = UTCNow().AddDate(0, 0, -1*int(days/24))
+	target.ResetID = resetID
+	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:
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrReplicationRemoteConnectionError, err), r.URL)
+		default:
+			writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
+		}
+	}
+	if err := startReplicationResync(ctx, bucket, arn, resetID, resetBeforeDate, objectAPI); err != nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
+			Bucket: bucket,
+			Err:    err,
+		}), r.URL)
+		return
+	}
+
+	data, err := json.Marshal(rinfo)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+	// Write success response.
+	writeSuccessResponseJSON(w, data)
+}
+
+// ResetBucketReplicationStatusHandler - returns the status of replication reset.
+// This API is a MinIO only extension
+func (api objectAPIHandlers) ResetBucketReplicationStatusHandler(w http.ResponseWriter, r *http.Request) {
+	ctx := newContext(r, w, "ResetBucketReplicationStatus")
+	defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
+
+	vars := mux.Vars(r)
+	bucket := vars["bucket"]
+	arn := r.URL.Query().Get("arn")
+	var err error
+
+	objectAPI := api.ObjectAPI()
+	if objectAPI == nil {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	if s3Error := checkRequestAuthType(ctx, r, policy.ResetBucketReplicationStateAction, bucket, ""); s3Error != ErrNone {
+		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL)
+		return
+	}
+
+	// Check if bucket exists.
+	if _, err := objectAPI.GetBucketInfo(ctx, bucket); 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
+	}
+
+	globalReplicationPool.resyncState.RLock()
+	brs, ok := globalReplicationPool.resyncState.statusMap[bucket]
+	globalReplicationPool.resyncState.RUnlock()
+	if !ok {
+		brs, err = loadBucketResyncMetadata(ctx, bucket, objectAPI)
+		if err != nil {
+			writeErrorResponse(ctx, w, errorCodes.ToAPIErrWithErr(ErrBadRequest, InvalidArgument{
+				Bucket: bucket,
+				Err:    fmt.Errorf("No replication resync status available for %s", arn),
+			}), r.URL)
+			return
+		}
+	}
+
+	var rinfo ResyncTargetsInfo
+	for tarn, st := range brs.TargetsMap {
+		if arn != "" && tarn != arn {
+			continue
+		}
+		rinfo.Targets = append(rinfo.Targets, ResyncTarget{
+			Arn:             tarn,
+			ResetID:         st.ResyncID,
+			StartTime:       st.StartTime,
+			EndTime:         st.EndTime,
+			ResyncStatus:    st.ResyncStatus.String(),
+			ReplicatedSize:  st.ReplicatedSize,
+			ReplicatedCount: st.ReplicatedCount,
+			FailedSize:      st.FailedSize,
+			FailedCount:     st.FailedCount,
+			Bucket:          st.Bucket,
+			Object:          st.Object,
+		})
+	}
+	data, err := json.Marshal(rinfo)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	// Write success response.
+	writeSuccessResponseJSON(w, data)
+}

cmd/bucket-replication-stats.go

@@ -135,6 +135,23 @@ func (r *ReplicationStats) GetInitialUsage(bucket string) BucketReplicationStats
 	return st.Clone()
 }
 
+// GetAll returns replication metrics for all buckets at once.
+func (r *ReplicationStats) GetAll() map[string]BucketReplicationStats {
+	if r == nil {
+		return map[string]BucketReplicationStats{}
+	}
+
+	r.RLock()
+	defer r.RUnlock()
+
+	bucketReplicationStats := make(map[string]BucketReplicationStats, len(r.Cache))
+	for k, v := range r.Cache {
+		bucketReplicationStats[k] = v.Clone()
+	}
+
+	return bucketReplicationStats
+}
+
 // Get replication metrics for a bucket from this node since this node came up.
 func (r *ReplicationStats) Get(bucket string) BucketReplicationStats {
 	if r == nil {
@@ -161,7 +178,7 @@ func NewReplicationStats(ctx context.Context, objectAPI ObjectLayer) *Replicatio
 
 // load replication metrics at cluster start from initial data usage
 func (r *ReplicationStats) loadInitialReplicationMetrics(ctx context.Context) {
-	rTimer := time.NewTimer(time.Minute * 1)
+	rTimer := time.NewTimer(time.Minute)
 	defer rTimer.Stop()
 	var (
 		dui DataUsageInfo
@@ -174,13 +191,12 @@ outer:
 			return
 		case <-rTimer.C:
 			dui, err = loadDataUsageFromBackend(GlobalContext, newObjectLayerFn())
-			if err != nil {
-				continue
-			}
 			// If LastUpdate is set, data usage is available.
-			if !dui.LastUpdate.IsZero() {
+			if err == nil && !dui.LastUpdate.IsZero() {
 				break outer
 			}
+
+			rTimer.Reset(time.Minute)
 		}
 	}
 

cmd/bucket-replication-utils.go

@@ -20,6 +20,7 @@ package cmd
 import (
 	"bytes"
 	"fmt"
+	"net/http"
 	"regexp"
 	"strconv"
 	"strings"
@@ -27,6 +28,7 @@ import (
 	"time"
 
 	"github.com/minio/minio/internal/bucket/replication"
+	xhttp "github.com/minio/minio/internal/http"
 )
 
 //go:generate msgp -file=$GOFILE
@@ -507,7 +509,7 @@ func getHealReplicateObjectInfo(objInfo ObjectInfo, rcfg replicationConfig) Repl
 				ObjectName: oi.Name,
 				VersionID:  oi.VersionID,
 			},
-		}, oi, ObjectOptions{}, nil)
+		}, oi, ObjectOptions{VersionSuspended: globalBucketVersioningSys.PrefixSuspended(oi.Bucket, oi.Name)}, nil)
 	} else {
 		dsc = mustReplicate(GlobalContext, oi.Bucket, oi.Name, getMustReplicateOptions(ObjectInfo{
 			UserDefined: oi.UserDefined,
@@ -699,3 +701,26 @@ func newBucketResyncStatus(bucket string) BucketReplicationResyncStatus {
 		Version:    resyncMetaVersion,
 	}
 }
+
+var contentRangeRegexp = regexp.MustCompile(`bytes ([0-9]+)-([0-9]+)/([0-9]+|\\*)`)
+
+// parse size from content-range header
+func parseSizeFromContentRange(h http.Header) (sz int64, err error) {
+	cr := h.Get(xhttp.ContentRange)
+	if cr == "" {
+		return sz, fmt.Errorf("Content-Range not set")
+	}
+	parts := contentRangeRegexp.FindStringSubmatch(cr)
+	if len(parts) != 4 {
+		return sz, fmt.Errorf("invalid Content-Range header %s", cr)
+	}
+	if parts[3] == "*" {
+		return -1, nil
+	}
+	var usz uint64
+	usz, err = strconv.ParseUint(parts[3], 10, 64)
+	if err != nil {
+		return sz, err
+	}
+	return int64(usz), nil
+}

cmd/bucket-replication.go

@@ -81,12 +81,13 @@ func getReplicationConfig(ctx context.Context, bucketName string) (rc *replicati
 		return rc, BucketReplicationConfigNotFound{Bucket: bucketName}
 	}
 
-	return globalBucketMetadataSys.GetReplicationConfig(ctx, bucketName)
+	rCfg, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucketName)
+	return rCfg, err
 }
 
 // validateReplicationDestination returns error if replication destination bucket missing or not configured
 // It also returns true if replication destination is same as this server.
-func validateReplicationDestination(ctx context.Context, bucket string, rCfg *replication.Config) (bool, APIError) {
+func validateReplicationDestination(ctx context.Context, bucket string, rCfg *replication.Config, checkRemote bool) (bool, APIError) {
 	var arns []string
 	if rCfg.RoleArn != "" {
 		arns = append(arns, rCfg.RoleArn)
@@ -95,26 +96,29 @@ func validateReplicationDestination(ctx context.Context, bucket string, rCfg *re
 			arns = append(arns, rule.Destination.String())
 		}
 	}
+	var sameTarget bool
 	for _, arnStr := range arns {
 		arn, err := madmin.ParseARN(arnStr)
 		if err != nil {
-			return false, errorCodes.ToAPIErrWithErr(ErrBucketRemoteArnInvalid, err)
+			return sameTarget, errorCodes.ToAPIErrWithErr(ErrBucketRemoteArnInvalid, err)
 		}
 		if arn.Type != madmin.ReplicationService {
-			return false, toAPIError(ctx, BucketRemoteArnTypeInvalid{Bucket: bucket})
+			return sameTarget, toAPIError(ctx, BucketRemoteArnTypeInvalid{Bucket: bucket})
 		}
 		clnt := globalBucketTargetSys.GetRemoteTargetClient(ctx, arnStr)
 		if clnt == nil {
-			return false, toAPIError(ctx, BucketRemoteTargetNotFound{Bucket: bucket})
+			return sameTarget, toAPIError(ctx, BucketRemoteTargetNotFound{Bucket: bucket})
 		}
+		if checkRemote { // validate remote bucket
 			if found, err := clnt.BucketExists(ctx, arn.Bucket); !found {
-			return false, errorCodes.ToAPIErrWithErr(ErrRemoteDestinationNotFoundError, err)
+				return sameTarget, errorCodes.ToAPIErrWithErr(ErrRemoteDestinationNotFoundError, err)
 			}
 			if ret, err := globalBucketObjectLockSys.Get(bucket); err == nil {
 				if ret.LockEnabled {
 					lock, _, _, _, err := clnt.GetObjectLockConfig(ctx, arn.Bucket)
 					if err != nil || lock != "Enabled" {
-					return false, errorCodes.ToAPIErrWithErr(ErrReplicationDestinationMissingLock, err)
+						return sameTarget, errorCodes.ToAPIErrWithErr(ErrReplicationDestinationMissingLock, err)
+					}
 				}
 			}
 		}
@@ -122,13 +126,20 @@ func validateReplicationDestination(ctx context.Context, bucket string, rCfg *re
 		c, ok := globalBucketTargetSys.arnRemotesMap[arnStr]
 		if ok {
 			if c.EndpointURL().String() == clnt.EndpointURL().String() {
-				sameTarget, _ := isLocalHost(clnt.EndpointURL().Hostname(), clnt.EndpointURL().Port(), globalMinioPort)
+				selfTarget, _ := isLocalHost(clnt.EndpointURL().Hostname(), clnt.EndpointURL().Port(), globalMinioPort)
-				return sameTarget, toAPIError(ctx, nil)
+				if !sameTarget {
+					sameTarget = selfTarget
+				}
+				continue
 			}
 		}
 	}
+
+	if len(arns) == 0 {
 		return false, toAPIError(ctx, BucketRemoteTargetNotFound{Bucket: bucket})
 	}
+	return sameTarget, toAPIError(ctx, nil)
+}
 
 type mustReplicateOptions struct {
 	meta               map[string]string
@@ -179,6 +190,17 @@ func mustReplicate(ctx context.Context, bucket, object string, mopts mustReplica
 		return
 	}
 
+	// object layer not initialized we return with no decision.
+	if newObjectLayerFn() == nil {
+		return
+	}
+
+	// Disable server-side replication on object prefixes which are excluded
+	// from versioning via the MinIO bucket versioning extension.
+	if globalBucketVersioningSys.PrefixSuspended(bucket, object) {
+		return
+	}
+
 	replStatus := mopts.ReplicationStatus()
 	if replStatus == replication.Replica && !mopts.isMetadataReplication() {
 		return
@@ -263,6 +285,11 @@ func checkReplicateDelete(ctx context.Context, bucket string, dobj ObjectToDelet
 	if delOpts.ReplicationRequest {
 		return
 	}
+	// Skip replication if this object's prefix is excluded from being
+	// versioned.
+	if !delOpts.Versioned {
+		return
+	}
 	opts := replication.ObjectOpts{
 		Name:         dobj.ObjectName,
 		SSEC:         crypto.SSEC.IsEncrypted(oi.UserDefined),
@@ -425,7 +452,7 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 		wg.Add(1)
 		go func(index int, tgt *TargetClient) {
 			defer wg.Done()
-			rinfo := replicateDeleteToTarget(ctx, dobj, objectAPI, tgt)
+			rinfo := replicateDeleteToTarget(ctx, dobj, tgt)
 			rinfos.Targets[index] = rinfo
 		}(idx, tgt)
 	}
@@ -456,8 +483,8 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 		VersionID:         versionID,
 		MTime:             dobj.DeleteMarkerMTime.Time,
 		DeleteReplication: drs,
-		Versioned:         globalBucketVersioningSys.Enabled(bucket),
+		Versioned:         globalBucketVersioningSys.PrefixEnabled(bucket, dobj.ObjectName),
-		VersionSuspended:  globalBucketVersioningSys.Suspended(bucket),
+		VersionSuspended:  globalBucketVersioningSys.PrefixSuspended(bucket, dobj.ObjectName),
 	})
 	if err != nil && !isErrVersionNotFound(err) { // VersionNotFound would be reported by pool that object version is missing on.
 		logger.LogIf(ctx, fmt.Errorf("Unable to update replication metadata for %s/%s(%s): %s", bucket, dobj.ObjectName, versionID, err))
@@ -482,7 +509,7 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 	}
 }
 
-func replicateDeleteToTarget(ctx context.Context, dobj DeletedObjectReplicationInfo, objectAPI ObjectLayer, tgt *TargetClient) (rinfo replicatedTargetInfo) {
+func replicateDeleteToTarget(ctx context.Context, dobj DeletedObjectReplicationInfo, tgt *TargetClient) (rinfo replicatedTargetInfo) {
 	versionID := dobj.DeleteMarkerVersionID
 	if versionID == "" {
 		versionID = dobj.VersionID
@@ -1018,8 +1045,13 @@ func replicateObjectToTarget(ctx context.Context, ri ReplicateObjectInfo, object
 		return
 	}
 
+	versioned := globalBucketVersioningSys.PrefixEnabled(bucket, object)
+	versionSuspended := globalBucketVersioningSys.PrefixSuspended(bucket, object)
+
 	gr, err = objectAPI.GetObjectNInfo(ctx, bucket, object, nil, http.Header{}, readLock, ObjectOptions{
 		VersionID:        objInfo.VersionID,
+		Versioned:        versioned,
+		VersionSuspended: versionSuspended,
 	})
 	if err != nil {
 		sendEvent(eventArgs{
@@ -1531,16 +1563,21 @@ func initBackgroundReplication(ctx context.Context, objectAPI ObjectLayer) {
 	go globalReplicationStats.loadInitialReplicationMetrics(ctx)
 }
 
+type proxyResult struct {
+	Proxy bool
+	Err   error
+}
+
 // get Reader from replication target if active-active replication is in place and
 // this node returns a 404
-func proxyGetToReplicationTarget(ctx context.Context, bucket, object string, rs *HTTPRangeSpec, h http.Header, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (gr *GetObjectReader, proxy bool) {
+func proxyGetToReplicationTarget(ctx context.Context, bucket, object string, rs *HTTPRangeSpec, h http.Header, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (gr *GetObjectReader, proxy proxyResult, err error) {
-	tgt, oi, proxy := proxyHeadToRepTarget(ctx, bucket, object, opts, proxyTargets)
+	tgt, oi, proxy := proxyHeadToRepTarget(ctx, bucket, object, rs, opts, proxyTargets)
-	if !proxy {
+	if !proxy.Proxy {
-		return nil, false
+		return nil, proxy, nil
 	}
-	fn, off, length, err := NewGetObjectReader(rs, oi, opts)
+	fn, _, _, err := NewGetObjectReader(nil, oi, opts)
 	if err != nil {
-		return nil, false
+		return nil, proxy, err
 	}
 	gopts := miniogo.GetObjectOptions{
 		VersionID:            opts.VersionID,
@@ -1548,33 +1585,47 @@ func proxyGetToReplicationTarget(ctx context.Context, bucket, object string, rs
 		Internal: miniogo.AdvancedGetOptions{
 			ReplicationProxyRequest: "true",
 		},
+		PartNumber: opts.PartNumber,
 	}
 	// get correct offsets for encrypted object
-	if off >= 0 && length >= 0 {
+	if rs != nil {
-		if err := gopts.SetRange(off, off+length-1); err != nil {
+		h, err := rs.ToHeader()
-			return nil, false
+		if err != nil {
+			return nil, proxy, err
 		}
+		gopts.Set(xhttp.Range, h)
 	}
 	// Make sure to match ETag when proxying.
 	if err = gopts.SetMatchETag(oi.ETag); err != nil {
-		return nil, false
+		return nil, proxy, err
 	}
 	c := miniogo.Core{Client: tgt.Client}
-	obj, _, _, err := c.GetObject(ctx, bucket, object, gopts)
+	obj, _, h, err := c.GetObject(ctx, bucket, object, gopts)
 	if err != nil {
-		return nil, false
+		return nil, proxy, err
 	}
 	closeReader := func() { obj.Close() }
-
 	reader, err := fn(obj, h, closeReader)
 	if err != nil {
-		return nil, false
+		return nil, proxy, err
 	}
 	reader.ObjInfo = oi.Clone()
-	return reader, true
+	if rs != nil {
+		contentSize, err := parseSizeFromContentRange(h)
+		if err != nil {
+			return nil, proxy, err
+		}
+		reader.ObjInfo.Size = contentSize
+	}
+
+	return reader, proxyResult{Proxy: true}, nil
+}
+
+func getProxyTargets(ctx context.Context, bucket, object string, opts ObjectOptions) (tgts *madmin.BucketTargets) {
+	if opts.VersionSuspended {
+		return &madmin.BucketTargets{}
 	}
 
-func getproxyTargets(ctx context.Context, bucket, object string, opts ObjectOptions) (tgts *madmin.BucketTargets) {
 	cfg, err := getReplicationConfig(ctx, bucket)
 	if err != nil || cfg == nil {
 		return &madmin.BucketTargets{}
@@ -1590,11 +1641,11 @@ func getproxyTargets(ctx context.Context, bucket, object string, opts ObjectOpti
 	return tgts
 }
 
-func proxyHeadToRepTarget(ctx context.Context, bucket, object string, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (tgt *TargetClient, oi ObjectInfo, proxy bool) {
+func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRangeSpec, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (tgt *TargetClient, oi ObjectInfo, proxy proxyResult) {
 	// this option is set when active-active replication is in place between site A -> B,
 	// and site B does not have the object yet.
 	if opts.ProxyRequest || (opts.ProxyHeaderSet && !opts.ProxyRequest) { // true only when site B sets MinIOSourceProxyRequest header
-		return nil, oi, false
+		return nil, oi, proxy
 	}
 	for _, t := range proxyTargets.Targets {
 		tgt = globalBucketTargetSys.GetRemoteTargetClient(ctx, t.Arn)
@@ -1612,9 +1663,22 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, opts Objec
 			Internal: miniogo.AdvancedGetOptions{
 				ReplicationProxyRequest: "true",
 			},
+			PartNumber: opts.PartNumber,
 		}
+		if rs != nil {
+			h, err := rs.ToHeader()
+			if err != nil {
+				logger.LogIf(ctx, fmt.Errorf("Invalid range header for %s/%s(%s) - %w", bucket, object, opts.VersionID, err))
+				continue
+			}
+			gopts.Set(xhttp.Range, h)
+		}
+
 		objInfo, err := tgt.StatObject(ctx, t.TargetBucket, object, gopts)
 		if err != nil {
+			if isErrInvalidRange(ErrorRespToObjectError(err, bucket, object)) {
+				return nil, oi, proxyResult{Err: err}
+			}
 			continue
 		}
 
@@ -1645,15 +1709,15 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, opts Objec
 		if ok {
 			oi.ContentEncoding = ce
 		}
-		return tgt, oi, true
+		return tgt, oi, proxyResult{Proxy: true}
 	}
-	return nil, oi, false
+	return nil, oi, proxy
 }
 
 // get object info from replication target if active-active replication is in place and
 // this node returns a 404
-func proxyHeadToReplicationTarget(ctx context.Context, bucket, object string, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (oi ObjectInfo, proxy bool) {
+func proxyHeadToReplicationTarget(ctx context.Context, bucket, object string, rs *HTTPRangeSpec, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (oi ObjectInfo, proxy proxyResult) {
-	_, oi, proxy = proxyHeadToRepTarget(ctx, bucket, object, opts, proxyTargets)
+	_, oi, proxy = proxyHeadToRepTarget(ctx, bucket, object, rs, opts, proxyTargets)
 	return oi, proxy
 }
 
@@ -1698,10 +1762,6 @@ func (c replicationConfig) Resync(ctx context.Context, oi ObjectInfo, dsc *Repli
 	if c.Empty() {
 		return
 	}
-	// existing object replication does not apply to un-versioned objects
-	if oi.VersionID == "" {
-		return
-	}
 
 	// Now overlay existing object replication choices for target
 	if oi.DeleteMarker {
@@ -1802,9 +1862,26 @@ func resyncTarget(oi ObjectInfo, arn string, resetID string, resetBeforeDate tim
 	return
 }
 
-// get the most current of in-memory replication stats  and data usage info from crawler.
+func getAllLatestReplicationStats(bucketsUsage map[string]BucketUsageInfo) (bucketsReplicationStats map[string]BucketReplicationStats) {
-func getLatestReplicationStats(bucket string, u BucketUsageInfo) (s BucketReplicationStats) {
+	peerBucketStatsList := globalNotificationSys.GetClusterAllBucketStats(GlobalContext)
-	bucketStats := globalNotificationSys.GetClusterBucketStats(GlobalContext, bucket)
+	bucketsReplicationStats = make(map[string]BucketReplicationStats, len(bucketsUsage))
+
+	for bucket, u := range bucketsUsage {
+		bucketStats := make([]BucketStats, len(peerBucketStatsList))
+		for i, peerBucketStats := range peerBucketStatsList {
+			bucketStat, ok := peerBucketStats[bucket]
+			if !ok {
+				continue
+			}
+			bucketStats[i] = bucketStat
+		}
+		bucketsReplicationStats[bucket] = calculateBucketReplicationStats(bucket, u, bucketStats)
+	}
+
+	return bucketsReplicationStats
+}
+
+func calculateBucketReplicationStats(bucket string, u BucketUsageInfo, bucketStats []BucketStats) (s BucketReplicationStats) {
 	// accumulate cluster bucket stats
 	stats := make(map[string]*BucketReplicationStat)
 	var totReplicaSize int64
@@ -1873,6 +1950,12 @@ func getLatestReplicationStats(bucket string, u BucketUsageInfo) (s BucketReplic
 	return s
 }
 
+// get the most current of in-memory replication stats  and data usage info from crawler.
+func getLatestReplicationStats(bucket string, u BucketUsageInfo) (s BucketReplicationStats) {
+	bucketStats := globalNotificationSys.GetClusterBucketStats(GlobalContext, bucket)
+	return calculateBucketReplicationStats(bucket, u, bucketStats)
+}
+
 const resyncTimeInterval = time.Minute * 10
 
 // periodicResyncMetaSave saves in-memory resync meta stats to disk in periodic intervals
@@ -1883,7 +1966,6 @@ func (p *ReplicationPool) periodicResyncMetaSave(ctx context.Context, objectAPI
 	for {
 		select {
 		case <-resyncTimer.C:
-			resyncTimer.Reset(resyncTimeInterval)
 			now := UTCNow()
 			p.resyncState.RLock()
 			for bucket, brs := range p.resyncState.statusMap {
@@ -1904,6 +1986,8 @@ func (p *ReplicationPool) periodicResyncMetaSave(ctx context.Context, objectAPI
 				}
 			}
 			p.resyncState.RUnlock()
+
+			resyncTimer.Reset(resyncTimeInterval)
 		case <-ctx.Done():
 			// server could be restarting - need
 			// to exit immediately
@@ -1923,6 +2007,7 @@ func resyncBucket(ctx context.Context, bucket, arn string, heal bool, objectAPI
 		st.EndTime = UTCNow()
 		st.ResyncStatus = resyncStatus
 		m.TargetsMap[arn] = st
+		globalReplicationPool.resyncState.statusMap[bucket] = m
 		globalReplicationPool.resyncState.Unlock()
 	}()
 	// Allocate new results channel to receive ObjectInfo.
@@ -1983,7 +2068,6 @@ func resyncBucket(ctx context.Context, bucket, arn string, heal bool, objectAPI
 		}
 
 		if roi.DeleteMarker || !roi.VersionPurgeStatus.Empty() {
-
 			versionID := ""
 			dmVersionID := ""
 			if roi.VersionPurgeStatus.Empty() {
@@ -2030,6 +2114,7 @@ func resyncBucket(ctx context.Context, bucket, arn string, heal bool, objectAPI
 			st.ReplicatedSize += roi.Size
 		}
 		m.TargetsMap[arn] = st
+		globalReplicationPool.resyncState.statusMap[bucket] = m
 		globalReplicationPool.resyncState.Unlock()
 	}
 	resyncStatus = ResyncCompleted
@@ -2110,10 +2195,6 @@ func (p *ReplicationPool) initResync(ctx context.Context, buckets []BucketInfo,
 	if objAPI == nil {
 		return errServerNotInitialized
 	}
-	// replication applies only to erasure coded setups
-	if !globalIsErasure {
-		return nil
-	}
 	// Load bucket metadata sys in background
 	go p.loadResync(ctx, buckets, objAPI)
 	return nil

cmd/bucket-stats.go

@@ -51,6 +51,9 @@ func (rl *ReplicationLatency) update(size int64, duration time.Duration) {
 	rl.UploadHistogram.Add(size, duration)
 }
 
+// BucketStatsMap captures bucket statistics for all buckets
+type BucketStatsMap map[string]BucketStats
+
 // BucketStats bucket statistics
 type BucketStats struct {
 	ReplicationStats BucketReplicationStats

cmd/bucket-stats_gen.go

@@ -792,6 +792,183 @@ func (z *BucketStats) Msgsize() (s int) {
 	return
 }
 
+// DecodeMsg implements msgp.Decodable
+func (z *BucketStatsMap) DecodeMsg(dc *msgp.Reader) (err error) {
+	var zb0003 uint32
+	zb0003, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	if (*z) == nil {
+		(*z) = make(BucketStatsMap, zb0003)
+	} else if len((*z)) > 0 {
+		for key := range *z {
+			delete((*z), key)
+		}
+	}
+	for zb0003 > 0 {
+		zb0003--
+		var zb0001 string
+		var zb0002 BucketStats
+		zb0001, err = dc.ReadString()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		var field []byte
+		_ = field
+		var zb0004 uint32
+		zb0004, err = dc.ReadMapHeader()
+		if err != nil {
+			err = msgp.WrapError(err, zb0001)
+			return
+		}
+		for zb0004 > 0 {
+			zb0004--
+			field, err = dc.ReadMapKeyPtr()
+			if err != nil {
+				err = msgp.WrapError(err, zb0001)
+				return
+			}
+			switch msgp.UnsafeString(field) {
+			case "ReplicationStats":
+				err = zb0002.ReplicationStats.DecodeMsg(dc)
+				if err != nil {
+					err = msgp.WrapError(err, zb0001, "ReplicationStats")
+					return
+				}
+			default:
+				err = dc.Skip()
+				if err != nil {
+					err = msgp.WrapError(err, zb0001)
+					return
+				}
+			}
+		}
+		(*z)[zb0001] = zb0002
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z BucketStatsMap) EncodeMsg(en *msgp.Writer) (err error) {
+	err = en.WriteMapHeader(uint32(len(z)))
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0005, zb0006 := range z {
+		err = en.WriteString(zb0005)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		// map header, size 1
+		// write "ReplicationStats"
+		err = en.Append(0x81, 0xb0, 0x52, 0x65, 0x70, 0x6c, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x53, 0x74, 0x61, 0x74, 0x73)
+		if err != nil {
+			return
+		}
+		err = zb0006.ReplicationStats.EncodeMsg(en)
+		if err != nil {
+			err = msgp.WrapError(err, zb0005, "ReplicationStats")
+			return
+		}
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z BucketStatsMap) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	o = msgp.AppendMapHeader(o, uint32(len(z)))
+	for zb0005, zb0006 := range z {
+		o = msgp.AppendString(o, zb0005)
+		// map header, size 1
+		// string "ReplicationStats"
+		o = append(o, 0x81, 0xb0, 0x52, 0x65, 0x70, 0x6c, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x53, 0x74, 0x61, 0x74, 0x73)
+		o, err = zb0006.ReplicationStats.MarshalMsg(o)
+		if err != nil {
+			err = msgp.WrapError(err, zb0005, "ReplicationStats")
+			return
+		}
+	}
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BucketStatsMap) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var zb0003 uint32
+	zb0003, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	if (*z) == nil {
+		(*z) = make(BucketStatsMap, zb0003)
+	} else if len((*z)) > 0 {
+		for key := range *z {
+			delete((*z), key)
+		}
+	}
+	for zb0003 > 0 {
+		var zb0001 string
+		var zb0002 BucketStats
+		zb0003--
+		zb0001, bts, err = msgp.ReadStringBytes(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		var field []byte
+		_ = field
+		var zb0004 uint32
+		zb0004, bts, err = msgp.ReadMapHeaderBytes(bts)
+		if err != nil {
+			err = msgp.WrapError(err, zb0001)
+			return
+		}
+		for zb0004 > 0 {
+			zb0004--
+			field, bts, err = msgp.ReadMapKeyZC(bts)
+			if err != nil {
+				err = msgp.WrapError(err, zb0001)
+				return
+			}
+			switch msgp.UnsafeString(field) {
+			case "ReplicationStats":
+				bts, err = zb0002.ReplicationStats.UnmarshalMsg(bts)
+				if err != nil {
+					err = msgp.WrapError(err, zb0001, "ReplicationStats")
+					return
+				}
+			default:
+				bts, err = msgp.Skip(bts)
+				if err != nil {
+					err = msgp.WrapError(err, zb0001)
+					return
+				}
+			}
+		}
+		(*z)[zb0001] = zb0002
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z BucketStatsMap) Msgsize() (s int) {
+	s = msgp.MapHeaderSize
+	if z != nil {
+		for zb0005, zb0006 := range z {
+			_ = zb0006
+			s += msgp.StringPrefixSize + len(zb0005) + 1 + 17 + zb0006.ReplicationStats.Msgsize()
+		}
+	}
+	return
+}
+
 // DecodeMsg implements msgp.Decodable
 func (z *ReplicationLatency) DecodeMsg(dc *msgp.Reader) (err error) {
 	var field []byte

cmd/bucket-stats_gen_test.go

@@ -348,6 +348,119 @@ func BenchmarkDecodeBucketStats(b *testing.B) {
 	}
 }
 
+func TestMarshalUnmarshalBucketStatsMap(t *testing.T) {
+	v := BucketStatsMap{}
+	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 BenchmarkMarshalMsgBucketStatsMap(b *testing.B) {
+	v := BucketStatsMap{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBucketStatsMap(b *testing.B) {
+	v := BucketStatsMap{}
+	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 BenchmarkUnmarshalBucketStatsMap(b *testing.B) {
+	v := BucketStatsMap{}
+	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 TestEncodeDecodeBucketStatsMap(t *testing.T) {
+	v := BucketStatsMap{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBucketStatsMap Msgsize() is inaccurate")
+	}
+
+	vn := BucketStatsMap{}
+	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 BenchmarkEncodeBucketStatsMap(b *testing.B) {
+	v := BucketStatsMap{}
+	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 BenchmarkDecodeBucketStatsMap(b *testing.B) {
+	v := BucketStatsMap{}
+	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 TestMarshalUnmarshalReplicationLatency(t *testing.T) {
 	v := ReplicationLatency{}
 	bts, err := v.MarshalMsg(nil)

cmd/bucket-targets.go

@@ -120,9 +120,6 @@ func (sys *BucketTargetSys) SetTarget(ctx context.Context, bucket string, tgt *m
 		return BucketRemoteConnectionErr{Bucket: tgt.TargetBucket, Err: err}
 	}
 	if tgt.Type == madmin.ReplicationService {
-		if !globalIsErasure {
-			return NotImplemented{Message: "Replication is not implemented in " + getMinioMode()}
-		}
 		if !globalBucketVersioningSys.Enabled(bucket) {
 			return BucketReplicationSourceNotVersioned{Bucket: bucket}
 		}
@@ -184,9 +181,6 @@ func (sys *BucketTargetSys) RemoveTarget(ctx context.Context, bucket, arnStr str
 	if globalIsGateway {
 		return nil
 	}
-	if !globalIsErasure {
-		return NotImplemented{Message: "Replication is not implemented in " + getMinioMode()}
-	}
 
 	if arnStr == "" {
 		return BucketRemoteArnInvalid{Bucket: bucket}
@@ -201,7 +195,7 @@ func (sys *BucketTargetSys) RemoveTarget(ctx context.Context, bucket, arnStr str
 		// reject removal of remote target if replication configuration is present
 		rcfg, err := getReplicationConfig(ctx, bucket)
 		if err == nil {
-			for _, tgtArn := range rcfg.FilterTargetArns(replication.ObjectOpts{}) {
+			for _, tgtArn := range rcfg.FilterTargetArns(replication.ObjectOpts{OpType: replication.AllReplicationType}) {
 				if err == nil && (tgtArn == arnStr || rcfg.RoleArn == arnStr) {
 					sys.RLock()
 					_, ok := sys.arnRemotesMap[arnStr]

cmd/bucket-versioning-handler.go

@@ -18,12 +18,14 @@
 package cmd
 
 import (
+	"encoding/base64"
 	"encoding/xml"
 	"io"
 	"net/http"
 
 	humanize "github.com/dustin/go-humanize"
 	"github.com/gorilla/mux"
+	"github.com/minio/madmin-go"
 	"github.com/minio/minio/internal/bucket/versioning"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/pkg/bucket/policy"
@@ -63,28 +65,28 @@ func (api objectAPIHandlers) PutBucketVersioningHandler(w http.ResponseWriter, r
 		return
 	}
 
-	if globalSiteReplicationSys.isEnabled() {
+	if globalSiteReplicationSys.isEnabled() && !v.Enabled() {
 		writeErrorResponse(ctx, w, APIError{
 			Code:           "InvalidBucketState",
-			Description:    "Cluster replication is enabled for this site, so the versioning state cannot be changed.",
+			Description:    "Cluster replication is enabled on this site, versioning cannot be suspended on bucket.",
-			HTTPStatusCode: http.StatusConflict,
+			HTTPStatusCode: http.StatusBadRequest,
 		}, r.URL)
 		return
 	}
 
-	if rcfg, _ := globalBucketObjectLockSys.Get(bucket); rcfg.LockEnabled && v.Suspended() {
+	if rcfg, _ := globalBucketObjectLockSys.Get(bucket); rcfg.LockEnabled && (v.Suspended() || v.PrefixesExcluded()) {
 		writeErrorResponse(ctx, w, APIError{
 			Code:           "InvalidBucketState",
-			Description:    "An Object Lock configuration is present on this bucket, so the versioning state cannot be changed.",
+			Description:    "An Object Lock configuration is present on this bucket, versioning cannot be suspended.",
-			HTTPStatusCode: http.StatusConflict,
+			HTTPStatusCode: http.StatusBadRequest,
 		}, r.URL)
 		return
 	}
 	if _, err := getReplicationConfig(ctx, bucket); err == nil && v.Suspended() {
 		writeErrorResponse(ctx, w, APIError{
 			Code:           "InvalidBucketState",
-			Description:    "A replication configuration is present on this bucket, so the versioning state cannot be changed.",
+			Description:    "A replication configuration is present on this bucket, bucket wide versioning cannot be suspended.",
-			HTTPStatusCode: http.StatusConflict,
+			HTTPStatusCode: http.StatusBadRequest,
 		}, r.URL)
 		return
 	}
@@ -100,6 +102,20 @@ func (api objectAPIHandlers) PutBucketVersioningHandler(w http.ResponseWriter, r
 		return
 	}
 
+	// 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.SRBucketMetaTypeVersionConfig,
+		Bucket:     bucket,
+		Versioning: &cfgStr,
+	}); err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
 	writeSuccessResponseHeadersOnly(w)
 }
 

cmd/bucket-versioning.go

@@ -17,44 +17,69 @@
 
 package cmd
 
-import "github.com/minio/minio/internal/bucket/versioning"
+import (
+	"strings"
+
+	"github.com/minio/minio/internal/bucket/versioning"
+	"github.com/minio/minio/internal/logger"
+)
 
 // BucketVersioningSys - policy subsystem.
 type BucketVersioningSys struct{}
 
 // Enabled enabled versioning?
 func (sys *BucketVersioningSys) Enabled(bucket string) bool {
-	vc, err := globalBucketMetadataSys.GetVersioningConfig(bucket)
+	vc, err := sys.Get(bucket)
 	if err != nil {
-		return false
+		logger.CriticalIf(GlobalContext, err)
 	}
 	return vc.Enabled()
 }
 
+// PrefixEnabled returns true is versioning is enabled at bucket level and if
+// the given prefix doesn't match any excluded prefixes pattern. This is
+// part of a MinIO versioning configuration extension.
+func (sys *BucketVersioningSys) PrefixEnabled(bucket, prefix string) bool {
+	vc, err := sys.Get(bucket)
+	if err != nil {
+		logger.CriticalIf(GlobalContext, err)
+	}
+	return vc.PrefixEnabled(prefix)
+}
+
 // Suspended suspended versioning?
 func (sys *BucketVersioningSys) Suspended(bucket string) bool {
-	vc, err := globalBucketMetadataSys.GetVersioningConfig(bucket)
+	vc, err := sys.Get(bucket)
 	if err != nil {
-		return false
+		logger.CriticalIf(GlobalContext, err)
 	}
 	return vc.Suspended()
 }
 
+// PrefixSuspended returns true if the given prefix matches an excluded prefix
+// pattern. This is part of a MinIO versioning configuration extension.
+func (sys *BucketVersioningSys) PrefixSuspended(bucket, prefix string) bool {
+	vc, err := sys.Get(bucket)
+	if err != nil {
+		logger.CriticalIf(GlobalContext, err)
+	}
+
+	return vc.PrefixSuspended(prefix)
+}
+
 // Get returns stored bucket policy
 func (sys *BucketVersioningSys) Get(bucket string) (*versioning.Versioning, error) {
 	if globalIsGateway {
-		objAPI := newObjectLayerFn()
+		// Gateway does not implement versioning.
-		if objAPI == nil {
+		return &versioning.Versioning{XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/"}, nil
-			return nil, errServerNotInitialized
-		}
-		return nil, NotImplemented{}
-	}
-	return globalBucketMetadataSys.GetVersioningConfig(bucket)
 	}
 
-// Reset BucketVersioningSys to initial state.
+	if bucket == minioMetaBucket || strings.HasPrefix(bucket, minioMetaBucket) {
-func (sys *BucketVersioningSys) Reset() {
+		return &versioning.Versioning{XMLNS: "http://s3.amazonaws.com/doc/2006-03-01/"}, nil
-	// There is currently no internal state.
+	}
+
+	vcfg, _, err := globalBucketMetadataSys.GetVersioningConfig(bucket)
+	return vcfg, err
 }
 
 // NewBucketVersioningSys - creates new versioning system.

cmd/callhome.go

@@ -0,0 +1,136 @@
+// 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"
+	"fmt"
+	"math/rand"
+	"time"
+
+	"github.com/minio/madmin-go"
+	"github.com/minio/minio/internal/logger"
+	uatomic "go.uber.org/atomic"
+)
+
+const (
+	// callhomeSchemaVersion1 is callhome schema version 1
+	callhomeSchemaVersion1 = "1"
+
+	// callhomeSchemaVersion is current callhome schema version.
+	callhomeSchemaVersion = callhomeSchemaVersion1
+
+	// callhomeCycleDefault is the default interval between two callhome cycles (24hrs)
+	callhomeCycleDefault = 24 * time.Hour
+)
+
+// CallhomeInfo - Contains callhome information
+type CallhomeInfo struct {
+	SchemaVersion string             `json:"schema_version"`
+	AdminInfo     madmin.InfoMessage `json:"admin_info"`
+}
+
+var (
+	enableCallhome            = uatomic.NewBool(false)
+	callhomeLeaderLockTimeout = newDynamicTimeout(30*time.Second, 10*time.Second)
+	callhomeFreq              = uatomic.NewDuration(callhomeCycleDefault)
+)
+
+func updateCallhomeParams(ctx context.Context, objAPI ObjectLayer) {
+	alreadyEnabled := enableCallhome.Load()
+	enableCallhome.Store(globalCallhomeConfig.Enable)
+	callhomeFreq.Store(globalCallhomeConfig.Frequency)
+
+	// If callhome was disabled earlier and has now been enabled,
+	// initialize the callhome process again.
+	if !alreadyEnabled && enableCallhome.Load() {
+		initCallhome(ctx, objAPI)
+	}
+}
+
+// initCallhome will start the callhome task in the background.
+func initCallhome(ctx context.Context, objAPI ObjectLayer) {
+	go func() {
+		r := rand.New(rand.NewSource(time.Now().UnixNano()))
+		// Leader node (that successfully acquires the lock inside runCallhome)
+		// will keep performing the callhome. If the leader goes down for some reason,
+		// the lock will be released and another node will acquire it and take over
+		// because of this loop.
+		for {
+			runCallhome(ctx, objAPI)
+			if !enableCallhome.Load() {
+				return
+			}
+
+			// callhome running on a different node.
+			// sleep for some time and try again.
+			duration := time.Duration(r.Float64() * float64(callhomeFreq.Load()))
+			if duration < time.Second {
+				// Make sure to sleep atleast a second to avoid high CPU ticks.
+				duration = time.Second
+			}
+			time.Sleep(duration)
+
+			if !enableCallhome.Load() {
+				return
+			}
+		}
+	}()
+}
+
+func runCallhome(ctx context.Context, objAPI ObjectLayer) {
+	// Make sure only 1 callhome is running on the cluster.
+	locker := objAPI.NewNSLock(minioMetaBucket, "callhome/runCallhome.lock")
+	lkctx, err := locker.GetLock(ctx, callhomeLeaderLockTimeout)
+	if err != nil {
+		return
+	}
+
+	ctx = lkctx.Context()
+	defer locker.Unlock(lkctx.Cancel)
+
+	callhomeTimer := time.NewTimer(callhomeFreq.Load())
+	defer callhomeTimer.Stop()
+
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case <-callhomeTimer.C:
+			if !enableCallhome.Load() {
+				// Stop the processing as callhome got disabled
+				return
+			}
+			performCallhome(ctx)
+
+			// Reset the timer for next cycle.
+			callhomeTimer.Reset(callhomeFreq.Load())
+		}
+	}
+}
+
+func performCallhome(ctx context.Context) {
+	err := sendCallhomeInfo(
+		CallhomeInfo{
+			SchemaVersion: callhomeSchemaVersion,
+			AdminInfo:     getServerInfo(ctx, nil),
+		})
+	if err != nil {
+		logger.LogIf(ctx, fmt.Errorf("Unable to perform callhome: %w", err))
+	}
+}

cmd/common-main.go

@@ -24,6 +24,7 @@ import (
 	"crypto/tls"
 	"crypto/x509"
 	"encoding/gob"
+	"encoding/pem"
 	"errors"
 	"fmt"
 	"io/ioutil"
@@ -32,6 +33,7 @@ import (
 	"net/http"
 	"net/url"
 	"os"
+	"path"
 	"path/filepath"
 	"runtime"
 	"sort"
@@ -100,12 +102,16 @@ func init() {
 		PersistOnFailure: false,
 	}
 
+	globalIsCICD = env.Get("MINIO_CI_CD", "") != "" || env.Get("CI", "") != ""
+
+	containers := IsKubernetes() || IsDocker() || IsBOSH() || IsDCOS() || IsPCFTile()
+
 	// Call to refresh will refresh names in cache. If you pass true, it will also
 	// remove cached names not looked up since the last call to Refresh. It is a good idea
 	// to call this method on a regular interval.
 	go func() {
 		var t *time.Ticker
-		if IsKubernetes() || IsDocker() || IsBOSH() || IsDCOS() || IsPCFTile() {
+		if containers {
 			t = time.NewTicker(1 * time.Minute)
 		} else {
 			t = time.NewTicker(10 * time.Minute)
@@ -172,11 +178,22 @@ func minioConfigToConsoleFeatures() {
 			os.Setenv("CONSOLE_LOG_QUERY_AUTH_TOKEN", value)
 		}
 	}
+	// pass the console subpath configuration
+	if value := env.Get(config.EnvMinIOBrowserRedirectURL, ""); value != "" {
+		subPath := path.Clean(pathJoin(strings.TrimSpace(globalBrowserRedirectURL.Path), SlashSeparator))
+		if subPath != SlashSeparator {
+			os.Setenv("CONSOLE_SUBPATH", subPath)
+		}
+	}
 	// Enable if prometheus URL is set.
 	if value := env.Get("MINIO_PROMETHEUS_URL", ""); value != "" {
 		os.Setenv("CONSOLE_PROMETHEUS_URL", value)
 		if value := env.Get("MINIO_PROMETHEUS_JOB_ID", "minio-job"); value != "" {
 			os.Setenv("CONSOLE_PROMETHEUS_JOB_ID", value)
+			// Support additional labels for more granular filtering.
+			if value := env.Get("MINIO_PROMETHEUS_EXTRA_LABELS", ""); value != "" {
+				os.Setenv("CONSOLE_PROMETHEUS_EXTRA_LABELS", value)
+			}
 		}
 	}
 	// Enable if LDAP is enabled.
@@ -184,23 +201,23 @@ func minioConfigToConsoleFeatures() {
 		os.Setenv("CONSOLE_LDAP_ENABLED", config.EnableOn)
 	}
 	// if IDP is enabled, set IDP environment variables
-	if globalOpenIDConfig.URL != nil {
+	if globalOpenIDConfig.ProviderCfgs[config.Default] != nil {
-		os.Setenv("CONSOLE_IDP_URL", globalOpenIDConfig.URL.String())
+		os.Setenv("CONSOLE_IDP_URL", globalOpenIDConfig.ProviderCfgs[config.Default].URL.String())
-		os.Setenv("CONSOLE_IDP_CLIENT_ID", globalOpenIDConfig.ClientID)
+		os.Setenv("CONSOLE_IDP_CLIENT_ID", globalOpenIDConfig.ProviderCfgs[config.Default].ClientID)
-		os.Setenv("CONSOLE_IDP_SECRET", globalOpenIDConfig.ClientSecret)
+		os.Setenv("CONSOLE_IDP_SECRET", globalOpenIDConfig.ProviderCfgs[config.Default].ClientSecret)
 		os.Setenv("CONSOLE_IDP_HMAC_SALT", globalDeploymentID)
-		os.Setenv("CONSOLE_IDP_HMAC_PASSPHRASE", globalOpenIDConfig.ClientID)
+		os.Setenv("CONSOLE_IDP_HMAC_PASSPHRASE", globalOpenIDConfig.ProviderCfgs[config.Default].ClientID)
-		os.Setenv("CONSOLE_IDP_SCOPES", strings.Join(globalOpenIDConfig.DiscoveryDoc.ScopesSupported, ","))
+		os.Setenv("CONSOLE_IDP_SCOPES", strings.Join(globalOpenIDConfig.ProviderCfgs[config.Default].DiscoveryDoc.ScopesSupported, ","))
-		if globalOpenIDConfig.ClaimUserinfo {
+		if globalOpenIDConfig.ProviderCfgs[config.Default].ClaimUserinfo {
 			os.Setenv("CONSOLE_IDP_USERINFO", config.EnableOn)
 		}
-		if globalOpenIDConfig.RedirectURIDynamic {
+		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.RedirectURI != "" {
+		if globalOpenIDConfig.ProviderCfgs[config.Default].RedirectURI != "" {
-			os.Setenv("CONSOLE_IDP_CALLBACK", globalOpenIDConfig.RedirectURI)
+			os.Setenv("CONSOLE_IDP_CALLBACK", globalOpenIDConfig.ProviderCfgs[config.Default].RedirectURI)
 		} else {
 			os.Setenv("CONSOLE_IDP_CALLBACK", getConsoleEndpoints()[0]+"/oauth_callback")
 		}
@@ -213,8 +230,8 @@ func minioConfigToConsoleFeatures() {
 	if globalSubnetConfig.APIKey != "" {
 		os.Setenv("CONSOLE_SUBNET_API_KEY", globalSubnetConfig.APIKey)
 	}
-	if globalSubnetConfig.Proxy != "" {
+	if globalSubnetConfig.ProxyURL != nil {
-		os.Setenv("CONSOLE_SUBNET_PROXY", globalSubnetConfig.Proxy)
+		os.Setenv("CONSOLE_SUBNET_PROXY", globalSubnetConfig.ProxyURL.String())
 	}
 }
 
@@ -327,7 +344,7 @@ func checkUpdate(mode string) {
 		return
 	}
 
-	logStartupMessage(prepareUpdateMessage("\nRun `mc admin update`", lrTime.Sub(crTime)))
+	logger.Info(prepareUpdateMessage("Run `mc admin update`", lrTime.Sub(crTime)))
 }
 
 func newConfigDirFromCtx(ctx *cli.Context, option string, getDefaultDir func() string) (*ConfigDir, bool) {
@@ -482,6 +499,13 @@ 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]
@@ -502,9 +526,6 @@ func parsEnvEntry(envEntry string) (envKV, error) {
 func minioEnvironFromFile(envConfigFile string) ([]envKV, error) {
 	f, err := os.Open(envConfigFile)
 	if err != nil {
-		if os.IsNotExist(err) { // ignore if file doesn't exist.
-			return nil, nil
-		}
 		return nil, err
 	}
 	defer f.Close()
@@ -600,7 +621,7 @@ 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 {
@@ -625,12 +646,11 @@ func handleCommonEnvVars() {
 			}
 			// Look for if URL has invalid values and return error.
 			if !((u.Scheme == "http" || u.Scheme == "https") &&
-				(u.Path == "/" || u.Path == "") && u.Opaque == "" &&
+				u.Opaque == "" &&
 				!u.ForceQuery && u.RawQuery == "" && u.Fragment == "") {
 				err := fmt.Errorf("URL contains unexpected resources, expected URL to be of http(s)://minio.example.com format: %v", u)
 				logger.Fatal(err, "Invalid MINIO_BROWSER_REDIRECT_URL value is environment variable")
 			}
-			u.Path = "" // remove any path component such as `/`
 			globalBrowserRedirectURL = u
 		}
 	}
@@ -664,8 +684,6 @@ func handleCommonEnvVars() {
 		globalRootDiskThreshold = size
 	}
 
-	globalIsCICD = env.Get("MINIO_CI_CD", "") != "" || env.Get("CI", "") != ""
-
 	domains := env.Get(config.EnvDomain, "")
 	if len(domains) != 0 {
 		for _, domainName := range strings.Split(domains, config.ValueSeparator) {
@@ -762,7 +780,7 @@ func handleCommonEnvVars() {
 				"         Please use %s and %s",
 				config.EnvAccessKey, config.EnvSecretKey,
 				config.EnvRootUser, config.EnvRootPassword)
-			logStartupMessage(color.RedBold(msg))
+			logger.Info(color.RedBold(msg))
 		}
 		globalActiveCred = cred
 	}
@@ -796,7 +814,29 @@ func handleCommonEnvVars() {
 				endpoints = append(endpoints, strings.Join(lbls, ""))
 			}
 		}
-		certificate, err := tls.LoadX509KeyPair(env.Get(config.EnvKESClientCert, ""), env.Get(config.EnvKESClientKey, ""))
+		// 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")
 		}
@@ -827,13 +867,6 @@ func handleCommonEnvVars() {
 	}
 }
 
-func logStartupMessage(msg string) {
-	if globalConsoleSys != nil {
-		globalConsoleSys.Send(msg, string(logger.All))
-	}
-	logger.StartupMessage(msg)
-}
-
 func getTLSConfig() (x509Certs []*x509.Certificate, manager *certs.Manager, secureConn bool, err error) {
 	if !(isFile(getPublicCertFile()) && isFile(getPrivateKeyFile())) {
 		return nil, nil, false, nil
@@ -910,6 +943,9 @@ func getTLSConfig() (x509Certs []*x509.Certificate, manager *certs.Manager, secu
 	}
 	secureConn = true
 
+	// Certs might be symlinks, reload them every 10 seconds.
+	manager.UpdateReloadDuration(10 * time.Second)
+
 	// syscall.SIGHUP to reload the certs.
 	manager.ReloadOnSignal(syscall.SIGHUP)
 

cmd/common-main_test.go

@@ -134,6 +134,24 @@ export MINIO_ROOT_PASSWORD=minio123`,
 			true,
 			nil,
 		},
+		{
+			`
+# MINIO_ROOT_USER=minioadmin
+# MINIO_ROOT_PASSWORD=minioadmin
+MINIO_ROOT_USER=minio
+MINIO_ROOT_PASSWORD=minio123`,
+			false,
+			[]envKV{
+				{
+					Key:   "MINIO_ROOT_USER",
+					Value: "minio",
+				},
+				{
+					Key:   "MINIO_ROOT_PASSWORD",
+					Value: "minio123",
+				},
+			},
+		},
 	}
 	for _, testCase := range testCases {
 		testCase := testCase
@@ -153,6 +171,11 @@ export MINIO_ROOT_PASSWORD=minio123`,
 			if err == nil && testCase.expectedErr {
 				t.Error(errors.New("expected error, found success"))
 			}
+
+			if len(ekvs) != len(testCase.expectedEkvs) {
+				t.Errorf("expected %v keys, got %v keys", len(testCase.expectedEkvs), len(ekvs))
+			}
+
 			if !reflect.DeepEqual(ekvs, testCase.expectedEkvs) {
 				t.Errorf("expected %v, got %v", testCase.expectedEkvs, ekvs)
 			}

cmd/config-current.go

@@ -28,15 +28,18 @@ import (
 	"github.com/minio/minio/internal/config"
 	"github.com/minio/minio/internal/config/api"
 	"github.com/minio/minio/internal/config/cache"
+	"github.com/minio/minio/internal/config/callhome"
 	"github.com/minio/minio/internal/config/compress"
 	"github.com/minio/minio/internal/config/dns"
 	"github.com/minio/minio/internal/config/etcd"
 	"github.com/minio/minio/internal/config/heal"
 	xldap "github.com/minio/minio/internal/config/identity/ldap"
 	"github.com/minio/minio/internal/config/identity/openid"
+	idplugin "github.com/minio/minio/internal/config/identity/plugin"
 	xtls "github.com/minio/minio/internal/config/identity/tls"
 	"github.com/minio/minio/internal/config/notify"
 	"github.com/minio/minio/internal/config/policy/opa"
+	polplugin "github.com/minio/minio/internal/config/policy/plugin"
 	"github.com/minio/minio/internal/config/scanner"
 	"github.com/minio/minio/internal/config/storageclass"
 	"github.com/minio/minio/internal/config/subnet"
@@ -55,7 +58,9 @@ func initHelp() {
 		config.IdentityLDAPSubSys:   xldap.DefaultKVS,
 		config.IdentityOpenIDSubSys: openid.DefaultKVS,
 		config.IdentityTLSSubSys:    xtls.DefaultKVS,
+		config.IdentityPluginSubSys: idplugin.DefaultKVS,
 		config.PolicyOPASubSys:      opa.DefaultKVS,
+		config.PolicyPluginSubSys:   polplugin.DefaultKVS,
 		config.SiteSubSys:           config.DefaultSiteKVS,
 		config.RegionSubSys:         config.DefaultRegionKVS,
 		config.APISubSys:            api.DefaultKVS,
@@ -66,6 +71,7 @@ func initHelp() {
 		config.HealSubSys:           heal.DefaultKVS,
 		config.ScannerSubSys:        scanner.DefaultKVS,
 		config.SubnetSubSys:         subnet.DefaultKVS,
+		config.CallhomeSubSys:       callhome.DefaultKVS,
 	}
 	for k, v := range notify.DefaultNotificationKVS {
 		kvs[k] = v
@@ -96,6 +102,7 @@ func initHelp() {
 		config.HelpKV{
 			Key:             config.IdentityOpenIDSubSys,
 			Description:     "enable OpenID SSO support",
+			MultipleTargets: true,
 		},
 		config.HelpKV{
 			Key:         config.IdentityLDAPSubSys,
@@ -106,8 +113,12 @@ func initHelp() {
 			Description: "enable X.509 TLS certificate SSO support",
 		},
 		config.HelpKV{
-			Key:         config.PolicyOPASubSys,
+			Key:         config.IdentityPluginSubSys,
-			Description: "[DEPRECATED] enable external OPA for policy enforcement",
+			Description: "enable Identity Plugin via external hook",
+		},
+		config.HelpKV{
+			Key:         config.PolicyPluginSubSys,
+			Description: "enable Access Management Plugin for policy enforcement",
 		},
 		config.HelpKV{
 			Key:         config.APISubSys,
@@ -192,6 +203,12 @@ func initHelp() {
 			Description: "set subnet config for the cluster e.g. api key",
 			Optional:    true,
 		},
+		config.HelpKV{
+			Key:         config.CallhomeSubSys,
+			Type:        "string",
+			Description: "enable callhome for the cluster",
+			Optional:    true,
+		},
 	}
 
 	if globalIsErasure {
@@ -217,7 +234,9 @@ func initHelp() {
 		config.IdentityOpenIDSubSys: openid.Help,
 		config.IdentityLDAPSubSys:   xldap.Help,
 		config.IdentityTLSSubSys:    xtls.Help,
+		config.IdentityPluginSubSys: idplugin.Help,
 		config.PolicyOPASubSys:      opa.Help,
+		config.PolicyPluginSubSys:   polplugin.Help,
 		config.LoggerWebhookSubSys:  logger.Help,
 		config.AuditWebhookSubSys:   logger.HelpWebhook,
 		config.AuditKafkaSubSys:     logger.HelpKafka,
@@ -232,6 +251,7 @@ func initHelp() {
 		config.NotifyWebhookSubSys:  notify.HelpWebhook,
 		config.NotifyESSubSys:       notify.HelpES,
 		config.SubnetSubSys:         subnet.HelpSubnet,
+		config.CallhomeSubSys:       callhome.HelpCallhome,
 	}
 
 	config.RegisterHelpSubSys(helpMap)
@@ -242,6 +262,10 @@ func initHelp() {
 			Key:         config.RegionSubSys,
 			Description: "[DEPRECATED - use `site` instead] label the location of the server",
 		},
+		config.PolicyOPASubSys: {
+			Key:         config.PolicyOPASubSys,
+			Description: "[DEPRECATED - use `policy_plugin` instead] enable external OPA for policy enforcement",
+		},
 	}
 
 	config.RegisterHelpDeprecatedSubSys(deprecatedHelpKVMap)
@@ -314,7 +338,7 @@ func validateSubSysConfig(s config.Config, subSys string, objAPI ObjectLayer) er
 			etcdClnt.Close()
 		}
 	case config.IdentityOpenIDSubSys:
-		if _, err := openid.LookupConfig(s[config.IdentityOpenIDSubSys][config.Default],
+		if _, err := openid.LookupConfig(s[config.IdentityOpenIDSubSys],
 			NewGatewayHTTPTransport(), xhttp.DrainBody, globalSite.Region); err != nil {
 			return err
 		}
@@ -334,15 +358,35 @@ func validateSubSysConfig(s config.Config, subSys string, objAPI ObjectLayer) er
 		if _, err := xtls.Lookup(s[config.IdentityTLSSubSys][config.Default]); err != nil {
 			return err
 		}
+	case config.IdentityPluginSubSys:
+		if _, err := idplugin.LookupConfig(s[config.IdentityPluginSubSys][config.Default],
+			NewGatewayHTTPTransport(), xhttp.DrainBody, globalSite.Region); err != nil {
+			return err
+		}
 	case config.SubnetSubSys:
 		if _, err := subnet.LookupConfig(s[config.SubnetSubSys][config.Default]); err != nil {
 			return err
 		}
+	case config.CallhomeSubSys:
+		if _, err := callhome.LookupConfig(s[config.CallhomeSubSys][config.Default]); err != nil {
+			return err
+		}
 	case config.PolicyOPASubSys:
+		// In case legacy OPA config is being set, we treat it as if the
+		// AuthZPlugin is being set.
+		subSys = config.PolicyPluginSubSys
+		fallthrough
+	case config.PolicyPluginSubSys:
+		if ppargs, err := polplugin.LookupConfig(s[config.PolicyPluginSubSys][config.Default],
+			NewGatewayHTTPTransport(), xhttp.DrainBody); err != nil {
+			return err
+		} else if ppargs.URL == nil {
+			// Check if legacy opa is configured.
 			if _, err := opa.LookupConfig(s[config.PolicyOPASubSys][config.Default],
 				NewGatewayHTTPTransport(), xhttp.DrainBody); err != nil {
 				return err
 			}
+		}
 	default:
 		if config.LoggerSubSystems.Contains(subSys) {
 			if err := logger.ValidateSubSysConfig(s, subSys); err != nil {
@@ -484,22 +528,6 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 		getRemoteInstanceTransport = newGatewayHTTPTransport(apiConfig.RemoteTransportDeadline)
 	})
 
-	if globalIsErasure && objAPI != nil {
-		setDriveCounts := objAPI.SetDriveCounts()
-		for i, setDriveCount := range setDriveCounts {
-			sc, err := storageclass.LookupConfig(s[config.StorageClassSubSys][config.Default], setDriveCount)
-			if err != nil {
-				logger.LogIf(ctx, fmt.Errorf("Unable to initialize storage class config: %w", err))
-				break
-			}
-			// if we validated all setDriveCounts and it was successful
-			// proceed to store the correct storage class globally.
-			if i == len(setDriveCounts)-1 {
-				globalStorageClass.Update(sc)
-			}
-		}
-	}
-
 	globalCacheConfig, err = cache.LookupConfig(s[config.CacheSubSys][config.Default])
 	if err != nil {
 		if globalIsGateway {
@@ -529,30 +557,48 @@ func lookupConfigs(s config.Config, objAPI ObjectLayer) {
 	}
 
 	if globalSTSTLSConfig.InsecureSkipVerify {
-		logger.Info("CRITICAL: enabling %s is not recommended in a production environment", xtls.EnvIdentityTLSSkipVerify)
+		logger.LogIf(ctx, fmt.Errorf("CRITICAL: enabling %s is not recommended in a production environment", xtls.EnvIdentityTLSSkipVerify))
 	}
 
-	globalOpenIDConfig, err = openid.LookupConfig(s[config.IdentityOpenIDSubSys][config.Default],
+	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))
 	}
 
-	opaCfg, err := opa.LookupConfig(s[config.PolicyOPASubSys][config.Default],
-		NewGatewayHTTPTransport(), xhttp.DrainBody)
-	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("Unable to initialize OPA: %w", err))
-	}
-
-	globalOpenIDValidators = getOpenIDValidators(globalOpenIDConfig)
-	globalPolicyOPA = opa.New(opaCfg)
-
 	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])
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("Unable to parse subnet configuration: %w", err))
@@ -616,7 +662,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 			return fmt.Errorf("Unable to apply scanner config: %w", err)
 		}
 		// update dynamic scanner values.
-		scannerCycle.Update(scannerCfg.Cycle)
+		scannerCycle.Store(scannerCfg.Cycle)
 		logger.LogIf(ctx, scannerSleeper.Update(scannerCfg.Delay, scannerCfg.MaxWait))
 	case config.LoggerWebhookSubSys:
 		loggerCfg, err := logger.LookupConfigForSubSys(s, config.LoggerWebhookSubSys)
@@ -626,7 +672,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 		userAgent := getUserAgent(getMinioMode())
 		for n, l := range loggerCfg.HTTP {
 			if l.Enabled {
-				l.LogOnce = logger.LogOnceIf
+				l.LogOnce = logger.LogOnceConsoleIf
 				l.UserAgent = userAgent
 				l.Transport = NewGatewayHTTPTransportWithClientCerts(l.ClientCert, l.ClientKey)
 				loggerCfg.HTTP[n] = l
@@ -644,7 +690,7 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 		userAgent := getUserAgent(getMinioMode())
 		for n, l := range loggerCfg.AuditWebhook {
 			if l.Enabled {
-				l.LogOnce = logger.LogOnceIf
+				l.LogOnce = logger.LogOnceConsoleIf
 				l.UserAgent = userAgent
 				l.Transport = NewGatewayHTTPTransportWithClientCerts(l.ClientCert, l.ClientKey)
 				loggerCfg.AuditWebhook[n] = l
@@ -670,6 +716,30 @@ func applyDynamicConfigForSubSys(ctx context.Context, objAPI ObjectLayer, s conf
 		if err != nil {
 			logger.LogIf(ctx, fmt.Errorf("Unable to update audit kafka targets: %w", err))
 		}
+	case config.StorageClassSubSys:
+		if globalIsErasure && objAPI != nil {
+			setDriveCounts := objAPI.SetDriveCounts()
+			for i, setDriveCount := range setDriveCounts {
+				sc, err := storageclass.LookupConfig(s[config.StorageClassSubSys][config.Default], setDriveCount)
+				if err != nil {
+					logger.LogIf(ctx, fmt.Errorf("Unable to initialize storage class config: %w", err))
+					break
+				}
+				// if we validated all setDriveCounts and it was successful
+				// proceed to store the correct storage class globally.
+				if i == len(setDriveCounts)-1 {
+					globalStorageClass.Update(sc)
+				}
+			}
+		}
+	case config.CallhomeSubSys:
+		callhomeCfg, err := callhome.LookupConfig(s[config.CallhomeSubSys][config.Default])
+		if err != nil {
+			logger.LogIf(ctx, fmt.Errorf("Unable to load callhome config: %w", err))
+		} else {
+			globalCallhomeConfig = callhomeCfg
+			updateCallhomeParams(ctx, objAPI)
+		}
 	}
 	globalServerConfigMu.Lock()
 	defer globalServerConfigMu.Unlock()
@@ -807,17 +877,3 @@ func loadConfig(objAPI ObjectLayer) error {
 
 	return nil
 }
-
-// getOpenIDValidators - returns ValidatorList which contains
-// enabled providers in server config.
-// A new authentication provider is added like below
-// * Add a new provider in pkg/iam/openid package.
-func getOpenIDValidators(cfg openid.Config) *openid.Validators {
-	validators := openid.NewValidators()
-
-	if cfg.Enabled {
-		validators.Add(&cfg)
-	}
-
-	return validators
-}

cmd/config-encrypted.go

@@ -74,7 +74,7 @@ func migrateIAMConfigsEtcdToEncrypted(ctx context.Context, client *etcd.Client)
 		if err != nil {
 			return err
 		}
-		logger.Info("Attempting to re-encrypt IAM users and policies on etcd with %q (%s)", stat.DefaultKey, stat.Name)
+		logger.Info(fmt.Sprintf("Attempting to re-encrypt IAM users and policies on etcd with %q (%s)", stat.DefaultKey, stat.Name))
 	}
 
 	listCtx, cancel := context.WithTimeout(ctx, 1*time.Minute)
@@ -143,7 +143,7 @@ func migrateConfigPrefixToEncrypted(objAPI ObjectLayer, encrypted bool) error {
 		if err != nil {
 			return err
 		}
-		logger.Info("Attempting to re-encrypt config, IAM users and policies on MinIO with %q (%s)", stat.DefaultKey, stat.Name)
+		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

cmd/config-migrate.go

@@ -2649,7 +2649,6 @@ func migrateV30ToV31MinioSys(objAPI ObjectLayer) error {
 
 	cfg.Version = "31"
 	cfg.OpenID = openid.Config{}
-	cfg.OpenID.JWKS.URL = &xnet.URL{}
 
 	cfg.Policy.OPA = opa.Args{
 		URL:       &xnet.URL{},

cmd/consolelogger.go

@@ -45,7 +45,7 @@ type HTTPConsoleLoggerSys struct {
 // NewConsoleLogger - creates new HTTPConsoleLoggerSys with all nodes subscribed to
 // the console logging pub sub system
 func NewConsoleLogger(ctx context.Context) *HTTPConsoleLoggerSys {
-	ps := pubsub.New()
+	ps := pubsub.New(8)
 	return &HTTPConsoleLoggerSys{
 		pubsub:  ps,
 		console: console.New(),
@@ -75,7 +75,7 @@ func (sys *HTTPConsoleLoggerSys) HasLogListeners() bool {
 }
 
 // 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) {
+func (sys *HTTPConsoleLoggerSys) Subscribe(subCh chan interface{}, doneCh <-chan struct{}, node string, last int, logKind string, filter func(entry interface{}) bool) error {
 	// Enable console logging for remote client.
 	if !sys.HasLogListeners() {
 		logger.AddSystemTarget(sys)
@@ -111,11 +111,12 @@ func (sys *HTTPConsoleLoggerSys) Subscribe(subCh chan interface{}, doneCh <-chan
 			select {
 			case subCh <- entry:
 			case <-doneCh:
-				return
+				return nil
 			}
 		}
 	}
-	sys.pubsub.Subscribe(subCh, doneCh, filter)
+
+	return sys.pubsub.Subscribe(subCh, doneCh, filter)
 }
 
 // Init if HTTPConsoleLoggerSys is valid, always returns nil right now

cmd/data-scanner.go

@@ -21,6 +21,7 @@ import (
 	"bytes"
 	"context"
 	"encoding/binary"
+	"encoding/json"
 	"errors"
 	"fmt"
 	"io/fs"
@@ -37,12 +38,14 @@ import (
 	"github.com/dustin/go-humanize"
 	"github.com/minio/madmin-go"
 	"github.com/minio/minio/internal/bucket/lifecycle"
+	"github.com/minio/minio/internal/bucket/object/lock"
 	"github.com/minio/minio/internal/bucket/replication"
 	"github.com/minio/minio/internal/color"
 	"github.com/minio/minio/internal/config/heal"
 	"github.com/minio/minio/internal/event"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/pkg/console"
+	uatomic "go.uber.org/atomic"
 )
 
 const (
@@ -64,9 +67,7 @@ var (
 	dataScannerLeaderLockTimeout = newDynamicTimeout(30*time.Second, 10*time.Second)
 	// Sleeper values are updated when config is loaded.
 	scannerSleeper = newDynamicSleeper(10, 10*time.Second)
-	scannerCycle   = &safeDuration{
+	scannerCycle   = uatomic.NewDuration(dataScannerStartDelay)
-		t: dataScannerStartDelay,
-	}
 )
 
 // initDataScanner will start the scanner in the background.
@@ -76,7 +77,7 @@ func initDataScanner(ctx context.Context, objAPI ObjectLayer) {
 		// Run the data scanner in a loop
 		for {
 			runDataScanner(ctx, objAPI)
-			duration := time.Duration(r.Float64() * float64(scannerCycle.Get()))
+			duration := time.Duration(r.Float64() * float64(scannerCycle.Load()))
 			if duration < time.Second {
 				// Make sure to sleep atleast a second to avoid high CPU ticks.
 				duration = time.Second
@@ -86,21 +87,61 @@ func initDataScanner(ctx context.Context, objAPI ObjectLayer) {
 	}()
 }
 
-type safeDuration struct {
+func getCycleScanMode(currentCycle, bitrotStartCycle uint64, bitrotStartTime time.Time) madmin.HealScanMode {
-	sync.Mutex
+	bitrotCycle := globalHealConfig.BitrotScanCycle()
-	t time.Duration
+	switch bitrotCycle {
+	case -1:
+		return madmin.HealNormalScan
+	case 0:
+		return madmin.HealDeepScan
 	}
 
-func (s *safeDuration) Update(t time.Duration) {
+	if currentCycle-bitrotStartCycle < healObjectSelectProb {
-	s.Lock()
+		return madmin.HealDeepScan
-	defer s.Unlock()
-	s.t = t
 	}
 
-func (s *safeDuration) Get() time.Duration {
+	if time.Since(bitrotStartTime) > bitrotCycle {
-	s.Lock()
+		return madmin.HealDeepScan
-	defer s.Unlock()
+	}
-	return s.t
+
+	return madmin.HealNormalScan
+}
+
+type backgroundHealInfo struct {
+	BitrotStartTime  time.Time           `json:"bitrotStartTime"`
+	BitrotStartCycle uint64              `json:"bitrotStartCycle"`
+	CurrentScanMode  madmin.HealScanMode `json:"currentScanMode"`
+}
+
+func readBackgroundHealInfo(ctx context.Context, objAPI ObjectLayer) backgroundHealInfo {
+	// Get last healing information
+	buf, err := readConfig(ctx, objAPI, backgroundHealInfoPath)
+	if err != nil {
+		if !errors.Is(err, errConfigNotFound) {
+			logger.LogIf(ctx, err)
+		}
+		return backgroundHealInfo{}
+	}
+	var info backgroundHealInfo
+	err = json.Unmarshal(buf, &info)
+	if err != nil {
+		logger.LogIf(ctx, err)
+		return backgroundHealInfo{}
+	}
+	return info
+}
+
+func saveBackgroundHealInfo(ctx context.Context, objAPI ObjectLayer, info backgroundHealInfo) {
+	b, err := json.Marshal(info)
+	if err != nil {
+		logger.LogIf(ctx, err)
+		return
+	}
+	// Get last healing information
+	err = saveConfig(ctx, objAPI, backgroundHealInfoPath, b)
+	if err != nil {
+		logger.LogIf(ctx, err)
+	}
 }
 
 // runDataScanner will start a data scanner.
@@ -130,7 +171,7 @@ func runDataScanner(pctx context.Context, objAPI ObjectLayer) {
 		}
 	}
 
-	scannerTimer := time.NewTimer(scannerCycle.Get())
+	scannerTimer := time.NewTimer(scannerCycle.Load())
 	defer scannerTimer.Stop()
 
 	for {
@@ -138,19 +179,28 @@ func runDataScanner(pctx context.Context, objAPI ObjectLayer) {
 		case <-ctx.Done():
 			return
 		case <-scannerTimer.C:
-			// Reset the timer for next cycle.
-			scannerTimer.Reset(scannerCycle.Get())
-
 			if intDataUpdateTracker.debug {
 				console.Debugln("starting scanner cycle")
 			}
 
+			bgHealInfo := readBackgroundHealInfo(ctx, objAPI)
+			scanMode := getCycleScanMode(nextBloomCycle, 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
+				}
+				saveBackgroundHealInfo(ctx, objAPI, newHealInfo)
+			}
+
 			// 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)
 			logger.LogIf(ctx, err)
-			err = objAPI.NSScanner(ctx, bf, results, uint32(nextBloomCycle))
+			err = objAPI.NSScanner(ctx, bf, results, uint32(nextBloomCycle), scanMode)
 			logger.LogIf(ctx, err)
 			if err == nil {
 				// Store new cycle...
@@ -161,6 +211,9 @@ func runDataScanner(pctx context.Context, objAPI ObjectLayer) {
 					logger.LogIf(ctx, err)
 				}
 			}
+
+			// Reset the timer for next cycle.
+			scannerTimer.Reset(scannerCycle.Load())
 		}
 	}
 }
@@ -182,6 +235,7 @@ type folderScanner struct {
 	dataUsageScannerDebug bool
 	healFolderInclude     uint32 // Include a clean folder one in n cycles.
 	healObjectSelect      uint32 // Do a heal check on an object once every n cycles. Must divide into healFolderInclude
+	scanMode              madmin.HealScanMode
 
 	disks       []StorageAPI
 	disksQuorum int
@@ -250,7 +304,7 @@ var globalScannerStats scannerStats
 // The returned cache will always be valid, but may not be updated from the existing.
 // 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) (dataUsageCache, error) {
+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: ")
@@ -279,6 +333,7 @@ func scanDataFolder(ctx context.Context, poolIdx, setIdx int, basePath string, c
 		dataUsageScannerDebug: intDataUpdateTracker.debug,
 		healFolderInclude:     0,
 		healObjectSelect:      0,
+		scanMode:              scanMode,
 		updates:               cache.Info.updates,
 	}
 
@@ -403,7 +458,7 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 		if filter != nil && ok && existing.Compacted {
 			// If folder isn't in filter and we have data, skip it completely.
 			if folder.name != dataUsageRoot && !filter.containsDir(folder.name) {
-				if f.healObjectSelect == 0 || !thisHash.mod(f.oldCache.Info.NextCycle, f.healFolderInclude/folder.objectHealProbDiv) {
+				if f.healObjectSelect == 0 || !thisHash.modAlt(f.oldCache.Info.NextCycle/folder.objectHealProbDiv, f.healFolderInclude/folder.objectHealProbDiv) {
 					f.newCache.copyWithChildren(&f.oldCache, thisHash, folder.parent)
 					f.updateCache.copyWithChildren(&f.oldCache, thisHash, folder.parent)
 					if f.dataUsageScannerDebug {
@@ -482,12 +537,15 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 				debug:       f.dataUsageScannerDebug,
 				lifeCycle:   activeLifeCycle,
 				replication: replicationCfg,
-				heal:        thisHash.mod(f.oldCache.Info.NextCycle, f.healObjectSelect/folder.objectHealProbDiv) && globalIsErasure,
 			}
+
+			item.heal.enabled = thisHash.modAlt(f.oldCache.Info.NextCycle/folder.objectHealProbDiv, f.healObjectSelect/folder.objectHealProbDiv) && globalIsErasure
+			item.heal.bitrot = f.scanMode == madmin.HealDeepScan
+
 			// if the drive belongs to an erasure set
 			// that is already being healed, skip the
 			// healing attempt on this drive.
-			item.heal = item.heal && f.healObjectSelect > 0
+			item.heal.enabled = item.heal.enabled && f.healObjectSelect > 0
 
 			sz, err := f.getSize(item)
 			if err != nil {
@@ -609,13 +667,13 @@ func (f *folderScanner) scanFolder(ctx context.Context, folder cachedFolder, int
 			// and the entry itself is compacted.
 			if !into.Compacted && f.oldCache.isCompacted(h) {
 				if !h.mod(f.oldCache.Info.NextCycle, dataUsageUpdateDirCycles) {
-					if f.healObjectSelect == 0 || !h.mod(f.oldCache.Info.NextCycle, f.healFolderInclude/folder.objectHealProbDiv) {
+					if f.healObjectSelect == 0 || !h.modAlt(f.oldCache.Info.NextCycle/folder.objectHealProbDiv, f.healFolderInclude/folder.objectHealProbDiv) {
 						// Transfer and add as child...
 						f.newCache.copyWithChildren(&f.oldCache, h, folder.parent)
 						into.addChild(h)
 						continue
 					}
-					folder.objectHealProbDiv = dataUsageUpdateDirCycles
+					folder.objectHealProbDiv = f.healFolderInclude
 				}
 			}
 			scanFolder(folder)
@@ -821,7 +879,10 @@ type scannerItem struct {
 	replication replicationConfig
 	lifeCycle   *lifecycle.Lifecycle
 	Typ         fs.FileMode
-	heal        bool // Has the object been selected for heal check?
+	heal        struct {
+		enabled bool
+		bitrot  bool
+	} // Has the object been selected for heal check?
 	debug bool
 }
 
@@ -874,9 +935,13 @@ func (i *scannerItem) applyHealing(ctx context.Context, o ObjectLayer, oi Object
 			console.Debugf(applyActionsLogPrefix+" heal checking: %v/%v\n", i.bucket, i.objectPath())
 		}
 	}
+	scanMode := madmin.HealNormalScan
+	if i.heal.bitrot {
+		scanMode = madmin.HealDeepScan
+	}
 	healOpts := madmin.HealOpts{
 		Remove:   healDeleteDangling,
-		ScanMode: globalHealConfig.ScanMode(),
+		ScanMode: scanMode,
 	}
 	res, err := o.HealObject(ctx, i.bucket, i.objectPath(), oi.VersionID, healOpts)
 	if err != nil && !errors.Is(err, NotImplemented{}) {
@@ -901,20 +966,8 @@ func (i *scannerItem) applyLifecycle(ctx context.Context, o ObjectLayer, oi Obje
 
 	atomic.AddUint64(&globalScannerStats.ilmChecks, 1)
 	versionID := oi.VersionID
-	action := i.lifeCycle.ComputeAction(
+	rCfg, _ := globalBucketObjectLockSys.Get(i.bucket)
-		lifecycle.ObjectOpts{
+	action := evalActionFromLifecycle(ctx, *i.lifeCycle, rCfg, oi, false)
-			Name:             i.objectPath(),
-			UserTags:         oi.UserTags,
-			ModTime:          oi.ModTime,
-			VersionID:        oi.VersionID,
-			DeleteMarker:     oi.DeleteMarker,
-			IsLatest:         oi.IsLatest,
-			NumVersions:      oi.NumVersions,
-			SuccessorModTime: oi.SuccessorModTime,
-			RestoreOngoing:   oi.RestoreOngoing,
-			RestoreExpires:   oi.RestoreExpires,
-			TransitionStatus: oi.TransitionedObject.Status,
-		})
 	if i.debug {
 		if versionID != "" {
 			console.Debugf(applyActionsLogPrefix+" lifecycle: %q (version-id=%s), Initial scan: %v\n", i.objectPath(), versionID, action)
@@ -986,9 +1039,13 @@ func (i *scannerItem) applyNewerNoncurrentVersionLimit(ctx context.Context, _ Ob
 	fivs = fivs[:lim+1]
 
 	rcfg, _ := globalBucketObjectLockSys.Get(i.bucket)
+	vcfg, _ := globalBucketVersioningSys.Get(i.bucket)
+
+	versioned := vcfg != nil && vcfg.Versioned(i.objectPath())
+
 	toDel := make([]ObjectToDelete, 0, len(overflowVersions))
 	for _, fi := range overflowVersions {
-		obj := fi.ToObjectInfo(i.bucket, i.objectPath())
+		obj := fi.ToObjectInfo(i.bucket, i.objectPath(), versioned)
 		// skip versions with object locking enabled
 		if rcfg.LockEnabled && enforceRetentionForDeletion(ctx, obj) {
 			if i.debug {
@@ -1040,7 +1097,7 @@ func (i *scannerItem) applyActions(ctx context.Context, o ObjectLayer, oi 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 {
+		if i.heal.enabled {
 			size = i.applyHealing(ctx, o, oi)
 		}
 		// replicate only if lifecycle rules are not applied.
@@ -1049,7 +1106,7 @@ func (i *scannerItem) applyActions(ctx context.Context, o ObjectLayer, oi Object
 	return size
 }
 
-func evalActionFromLifecycle(ctx context.Context, lc lifecycle.Lifecycle, obj ObjectInfo, debug bool) (action lifecycle.Action) {
+func evalActionFromLifecycle(ctx context.Context, lc lifecycle.Lifecycle, lr lock.Retention, obj ObjectInfo, debug bool) (action lifecycle.Action) {
 	action = lc.ComputeAction(obj.ToLifecycleOpts())
 	if debug {
 		console.Debugf(applyActionsLogPrefix+" lifecycle: Secondary scan: %v\n", action)
@@ -1065,9 +1122,7 @@ func evalActionFromLifecycle(ctx context.Context, lc lifecycle.Lifecycle, obj Ob
 		if obj.VersionID == "" {
 			return lifecycle.NoneAction
 		}
-		if rcfg, _ := globalBucketObjectLockSys.Get(obj.Bucket); rcfg.LockEnabled {
+		if lr.LockEnabled && enforceRetentionForDeletion(ctx, obj) {
-			locked := enforceRetentionForDeletion(ctx, obj)
-			if locked {
 			if debug {
 				if obj.VersionID != "" {
 					console.Debugf(applyActionsLogPrefix+" lifecycle: %s v(%s) is locked, not deleting\n", obj.Name, obj.VersionID)
@@ -1078,7 +1133,6 @@ func evalActionFromLifecycle(ctx context.Context, lc lifecycle.Lifecycle, obj Ob
 			return lifecycle.NoneAction
 		}
 	}
-	}
 
 	return action
 }
@@ -1116,7 +1170,8 @@ func applyExpiryOnNonTransitionedObjects(ctx context.Context, objLayer ObjectLay
 		opts.VersionID = obj.VersionID
 	}
 	if opts.VersionID == "" {
-		opts.Versioned = globalBucketVersioningSys.Enabled(obj.Bucket)
+		opts.Versioned = globalBucketVersioningSys.PrefixEnabled(obj.Bucket, obj.Name)
+		opts.VersionSuspended = globalBucketVersioningSys.PrefixSuspended(obj.Bucket, obj.Name)
 	}
 
 	obj, err := objLayer.DeleteObject(ctx, obj.Bucket, obj.Name, opts)
@@ -1175,6 +1230,9 @@ 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() {
+		return
+	}
 
 	if oi.DeleteMarker || !oi.VersionPurgeStatus.Empty() {
 		// heal delete marker replication failure or versioned delete replication failure
@@ -1195,10 +1253,10 @@ func (i *scannerItem) healReplication(ctx context.Context, o ObjectLayer, oi Obj
 		roi.OpType = replication.ExistingObjectReplicationType
 	}
 
-	if roi.TargetStatuses != nil {
+	if sizeS.replTargetStats == nil && len(roi.TargetStatuses) > 0 {
-		if sizeS.replTargetStats == nil {
 		sizeS.replTargetStats = make(map[string]replTargetSizeSummary)
 	}
+
 	for arn, tgtStatus := range roi.TargetStatuses {
 		tgtSizeS, ok := sizeS.replTargetStats[arn]
 		if !ok {
@@ -1215,13 +1273,12 @@ func (i *scannerItem) healReplication(ctx context.Context, o ObjectLayer, oi Obj
 			tgtSizeS.failedCount++
 			sizeS.failedSize += oi.Size
 			sizeS.failedCount++
-			case replication.Completed, "COMPLETE":
+		case replication.Completed, replication.CompletedLegacy:
 			tgtSizeS.replicatedSize += oi.Size
 			sizeS.replicatedSize += oi.Size
 		}
 		sizeS.replTargetStats[arn] = tgtSizeS
 	}
-	}
 
 	switch oi.ReplicationStatus {
 	case replication.Pending, replication.Failed:

cmd/data-usage-cache.go

@@ -369,6 +369,17 @@ func (h dataUsageHash) mod(cycle uint32, cycles uint32) bool {
 	return uint32(xxhash.Sum64String(string(h)))%cycles == cycle%cycles
 }
 
+// modAlt returns true if the hash mod cycles == cycle.
+// This is out of sync with mod.
+// If cycles is 0 false is always returned.
+// If cycles is 1 true is always returned (as expected).
+func (h dataUsageHash) modAlt(cycle uint32, cycles uint32) bool {
+	if cycles <= 1 {
+		return cycles == 1
+	}
+	return uint32(xxhash.Sum64String(string(h))>>32)%(cycles) == cycle%cycles
+}
+
 // addChild will add a child based on its hash.
 // If it already exists it will not be added again.
 func (e *dataUsageEntry) addChild(hash dataUsageHash) {
@@ -526,6 +537,7 @@ func (d *dataUsageCache) dui(path string, buckets []BucketInfo) DataUsageInfo {
 	dui := DataUsageInfo{
 		LastUpdate:         d.Info.LastUpdate,
 		ObjectsTotalCount:  flat.Objects,
+		VersionsTotalCount: flat.Versions,
 		ObjectsTotalSize:   uint64(flat.Size),
 		BucketsCount:       uint64(len(e.Children)),
 		BucketsUsage:       d.bucketsUsageInfo(buckets),
@@ -777,6 +789,7 @@ func (d *dataUsageCache) bucketsUsageInfo(buckets []BucketInfo) map[string]Bucke
 		flat := d.flatten(*e)
 		bui := BucketUsageInfo{
 			Size:                 uint64(flat.Size),
+			VersionsCount:        flat.Versions,
 			ObjectsCount:         flat.Objects,
 			ObjectSizesHistogram: flat.ObjSizes.toMap(),
 		}

cmd/data-usage-utils.go

@@ -60,6 +60,7 @@ type BucketUsageInfo struct {
 
 	ObjectsCount         uint64                           `json:"objectsCount"`
 	ObjectSizesHistogram map[string]uint64                `json:"objectsSizesHistogram"`
+	VersionsCount        uint64                           `json:"versionsCount"`
 	ReplicaSize          uint64                           `json:"objectReplicaTotalSize"`
 	ReplicationInfo      map[string]BucketTargetUsageInfo `json:"objectsReplicationInfo"`
 }
@@ -73,6 +74,9 @@ type DataUsageInfo struct {
 	// Objects total count across all buckets
 	ObjectsTotalCount uint64 `json:"objectsCount"`
 
+	// Objects total count across all buckets
+	VersionsTotalCount uint64 `json:"versionsCount"`
+
 	// Objects total size across all buckets
 	ObjectsTotalSize uint64                           `json:"objectsTotalSize"`
 	ReplicationInfo  map[string]BucketTargetUsageInfo `json:"objectsReplicationInfo"`
@@ -92,32 +96,37 @@ type DataUsageInfo struct {
 }
 
 func (dui DataUsageInfo) tierStats() []madmin.TierInfo {
-	if globalTierConfigMgr.Empty() {
+	if dui.TierStats == nil {
 		return nil
 	}
 
-	ts := make(map[string]madmin.TierStats)
+	cfgs := globalTierConfigMgr.ListTiers()
-	// Add configured remote tiers
+	if len(cfgs) == 0 {
-	for tier := range globalTierConfigMgr.Tiers {
+		return nil
-		ts[tier] = madmin.TierStats{}
 	}
+
+	ts := make(map[string]madmin.TierStats, len(cfgs)+1)
+	infos := make([]madmin.TierInfo, 0, len(ts))
+
 	// Add STANDARD (hot-tier)
 	ts[minioHotTier] = madmin.TierStats{}
+	infos = append(infos, madmin.TierInfo{
+		Name: minioHotTier,
+		Type: "internal",
+	})
+	// Add configured remote tiers
+	for _, cfg := range cfgs {
+		ts[cfg.Name] = madmin.TierStats{}
+		infos = append(infos, madmin.TierInfo{
+			Name: cfg.Name,
+			Type: cfg.Type.String(),
+		})
+	}
 
 	ts = dui.TierStats.adminStats(ts)
-	infos := make([]madmin.TierInfo, 0, len(ts))
+	for i := range infos {
-	for tier, st := range ts {
+		info := infos[i]
-		var tierType string
+		infos[i].Stats = ts[info.Name]
-		if tier == minioHotTier {
-			tierType = "internal"
-		} else {
-			tierType = globalTierConfigMgr.Tiers[tier].Type.String()
-		}
-		infos = append(infos, madmin.TierInfo{
-			Name:  tier,
-			Type:  tierType,
-			Stats: st,
-		})
 	}
 
 	sort.Slice(infos, func(i, j int) bool {
@@ -131,3 +140,31 @@ func (dui DataUsageInfo) tierStats() []madmin.TierInfo {
 	})
 	return infos
 }
+
+func (dui DataUsageInfo) tierMetrics() (metrics []Metric) {
+	if dui.TierStats == nil {
+		return nil
+	}
+	// e.g minio_cluster_ilm_transitioned_bytes{tier="S3TIER-1"}=136314880
+	//     minio_cluster_ilm_transitioned_objects{tier="S3TIER-1"}=1
+	//     minio_cluster_ilm_transitioned_versions{tier="S3TIER-1"}=3
+	for tier, st := range dui.TierStats.Tiers {
+		metrics = append(metrics, Metric{
+			Description:    getClusterTransitionedBytesMD(),
+			Value:          float64(st.TotalSize),
+			VariableLabels: map[string]string{"tier": tier},
+		})
+		metrics = append(metrics, Metric{
+			Description:    getClusterTransitionedObjectsMD(),
+			Value:          float64(st.NumObjects),
+			VariableLabels: map[string]string{"tier": tier},
+		})
+		metrics = append(metrics, Metric{
+			Description:    getClusterTransitionedVersionsMD(),
+			Value:          float64(st.NumVersions),
+			VariableLabels: map[string]string{"tier": tier},
+		})
+	}
+
+	return metrics
+}

cmd/data-usage.go

@@ -35,6 +35,8 @@ const (
 	dataUsageBloomName     = ".bloomcycle.bin"
 	dataUsageBloomNamePath = bucketMetaPrefix + SlashSeparator + dataUsageBloomName
 
+	backgroundHealInfoPath = bucketMetaPrefix + SlashSeparator + ".background-heal.json"
+
 	dataUsageCacheName = ".usage-cache.bin"
 )
 

cmd/data-usage_test.go

@@ -67,7 +67,7 @@ func TestDataUsageUpdate(t *testing.T) {
 		return
 	}
 
-	got, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: bucket}}, getSize)
+	got, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: bucket}}, getSize, 0)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -178,7 +178,7 @@ func TestDataUsageUpdate(t *testing.T) {
 	}
 	// Changed dir must be picked up in this many cycles.
 	for i := 0; i < dataUsageUpdateDirCycles; i++ {
-		got, err = scanDataFolder(context.Background(), 0, 0, base, got, getSize)
+		got, err = scanDataFolder(context.Background(), 0, 0, base, got, getSize, 0)
 		got.Info.NextCycle++
 		if err != nil {
 			t.Fatal(err)
@@ -289,7 +289,7 @@ func TestDataUsageUpdatePrefix(t *testing.T) {
 		}
 		return
 	}
-	got, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: "bucket"}}, getSize)
+	got, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: "bucket"}}, getSize, 0)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -423,7 +423,7 @@ func TestDataUsageUpdatePrefix(t *testing.T) {
 	}
 	// Changed dir must be picked up in this many cycles.
 	for i := 0; i < dataUsageUpdateDirCycles; i++ {
-		got, err = scanDataFolder(context.Background(), 0, 0, base, got, getSize)
+		got, err = scanDataFolder(context.Background(), 0, 0, base, got, getSize, 0)
 		got.Info.NextCycle++
 		if err != nil {
 			t.Fatal(err)
@@ -575,7 +575,7 @@ func TestDataUsageCacheSerialize(t *testing.T) {
 		}
 		return
 	}
-	want, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: bucket}}, getSize)
+	want, err := scanDataFolder(context.Background(), 0, 0, base, dataUsageCache{Info: dataUsageCacheInfo{Name: bucket}}, getSize, 0)
 	if err != nil {
 		t.Fatal(err)
 	}

cmd/disk-cache-backend.go

@@ -24,6 +24,7 @@ import (
 	"crypto/rand"
 	"encoding/base64"
 	"encoding/hex"
+	"errors"
 	"fmt"
 	"io"
 	"io/ioutil"
@@ -252,16 +253,17 @@ func (c *diskCache) diskUsageLow() bool {
 // Returns if the disk usage reaches  or exceeds configured cache quota when size is added.
 // If current usage without size exceeds high watermark a GC is automatically queued.
 func (c *diskCache) diskSpaceAvailable(size int64) bool {
+	reqInfo := (&logger.ReqInfo{}).AppendTags("cachePath", c.dir)
+	ctx := logger.SetReqInfo(GlobalContext, reqInfo)
+
 	gcTriggerPct := c.quotaPct * c.highWatermark / 100
 	di, err := disk.GetInfo(c.dir)
 	if err != nil {
-		reqInfo := (&logger.ReqInfo{}).AppendTags("cachePath", c.dir)
-		ctx := logger.SetReqInfo(GlobalContext, reqInfo)
 		logger.LogIf(ctx, err)
 		return false
 	}
 	if di.Total == 0 {
-		logger.Info("diskCache: Received 0 total disk size")
+		logger.LogIf(ctx, errors.New("diskCache: Received 0 total disk size"))
 		return false
 	}
 	usedPercent := float64(di.Used) * 100 / float64(di.Total)
@@ -882,6 +884,7 @@ func (c *diskCache) put(ctx context.Context, bucket, object string, data io.Read
 	}
 
 	if err := os.MkdirAll(cachePath, 0o777); err != nil {
+		removeAll(cachePath)
 		return oi, err
 	}
 	metadata := cloneMSS(opts.UserDefined)
@@ -890,6 +893,7 @@ func (c *diskCache) put(ctx context.Context, bucket, object string, data io.Read
 	if globalCacheKMS != nil {
 		reader, err = newCacheEncryptReader(data, bucket, object, metadata)
 		if err != nil {
+			removeAll(cachePath)
 			return oi, err
 		}
 		actualSize, _ = sio.EncryptedSize(uint64(size))
@@ -1626,8 +1630,6 @@ func (c *diskCache) cleanupStaleUploads(ctx context.Context) {
 		case <-ctx.Done():
 			return
 		case <-timer.C:
-			// Reset for the next interval
-			timer.Reset(cacheStaleUploadCleanupInterval)
 			now := time.Now()
 			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 {
@@ -1658,6 +1660,9 @@ func (c *diskCache) cleanupStaleUploads(ctx context.Context) {
 				}
 				return nil
 			})
+
+			// Reset for the next interval
+			timer.Reset(cacheStaleUploadCleanupInterval)
 		}
 	}
 }

cmd/disk-cache-utils.go

@@ -19,19 +19,17 @@ package cmd
 
 import (
 	"container/list"
-	"encoding/hex"
 	"errors"
 	"fmt"
 	"io"
 	"math"
 	"os"
-	"path"
 	"strconv"
 	"strings"
 	"time"
 
 	"github.com/minio/minio/internal/crypto"
-	"github.com/minio/minio/internal/kms"
+	"github.com/minio/minio/internal/etag"
 )
 
 // CacheStatusType - whether the request was served from cache.
@@ -234,77 +232,85 @@ func isCacheEncrypted(meta map[string]string) bool {
 
 // decryptCacheObjectETag tries to decrypt the ETag saved in encrypted format using the cache KMS
 func decryptCacheObjectETag(info *ObjectInfo) error {
-	// Directories are never encrypted.
 	if info.IsDir {
-		return nil
+		return nil // Directories are never encrypted.
 	}
-	encrypted := crypto.S3.IsEncrypted(info.UserDefined) && isCacheEncrypted(info.UserDefined)
 
-	switch {
+	// Depending on the SSE type we handle ETags slightly
-	case encrypted:
+	// differently. ETags encrypted with SSE-S3 must be
-		if globalCacheKMS == nil {
+	// decrypted first, since the client expects that
-			return errKMSNotConfigured
+	// a single-part SSE-S3 ETag is equal to the content MD5.
-		}
+	//
-		if len(info.Parts) > 0 { // multipart ETag is not encrypted since it is not md5sum
+	// For all other SSE types, the ETag is not the content MD5.
-			return nil
+	// Therefore, we don't decrypt but only format it.
-		}
+	switch kind, ok := crypto.IsEncrypted(info.UserDefined); {
-		keyID, kmsKey, sealedKey, err := crypto.S3.ParseMetadata(info.UserDefined)
+	case ok && kind == crypto.S3 && isCacheEncrypted(info.UserDefined):
+		ETag, err := etag.Parse(info.ETag)
 		if err != nil {
 			return err
 		}
-		extKey, err := globalCacheKMS.DecryptKey(keyID, kmsKey, kms.Context{info.Bucket: path.Join(info.Bucket, info.Name)})
+		if !ETag.IsEncrypted() {
-		if err != nil {
+			info.ETag = ETag.Format().String()
-			return err
-		}
-		var objectKey crypto.ObjectKey
-		if err = objectKey.Unseal(extKey, sealedKey, crypto.S3.String(), info.Bucket, info.Name); err != nil {
-			return err
-		}
-		etagStr := tryDecryptETag(objectKey[:], info.ETag, false)
-		// backend ETag was hex encoded before encrypting, so hex decode to get actual ETag
-		etag, err := hex.DecodeString(etagStr)
-		if err != nil {
-			return err
-		}
-		info.ETag = string(etag)
 			return nil
 		}
 
+		key, err := crypto.S3.UnsealObjectKey(globalCacheKMS, info.UserDefined, info.Bucket, info.Name)
+		if err != nil {
+			return err
+		}
+		ETag, err = etag.Decrypt(key[:], ETag)
+		if err != nil {
+			return err
+		}
+		info.ETag = ETag.Format().String()
+	case ok && (kind == crypto.S3KMS || kind == crypto.SSEC) && isCacheEncrypted(info.UserDefined):
+		ETag, err := etag.Parse(info.ETag)
+		if err != nil {
+			return err
+		}
+		info.ETag = ETag.Format().String()
+	}
 	return nil
 }
 
 // decryptCacheObjectETag tries to decrypt the ETag saved in encrypted format using the cache KMS
 func decryptCachePartETags(c *cacheMeta) ([]string, error) {
-	var partETags []string
+	// Depending on the SSE type we handle ETags slightly
-	encrypted := crypto.S3.IsEncrypted(c.Meta) && isCacheEncrypted(c.Meta)
+	// differently. ETags encrypted with SSE-S3 must be
-
+	// decrypted first, since the client expects that
-	switch {
+	// a single-part SSE-S3 ETag is equal to the content MD5.
-	case encrypted:
+	//
-		if globalCacheKMS == nil {
+	// For all other SSE types, the ETag is not the content MD5.
-			return partETags, errKMSNotConfigured
+	// Therefore, we don't decrypt but only format it.
-		}
+	switch kind, ok := crypto.IsEncrypted(c.Meta); {
-		keyID, kmsKey, sealedKey, err := crypto.S3.ParseMetadata(c.Meta)
+	case ok && kind == crypto.S3 && isCacheEncrypted(c.Meta):
+		key, err := crypto.S3.UnsealObjectKey(globalCacheKMS, c.Meta, c.Bucket, c.Object)
 		if err != nil {
-			return partETags, err
+			return nil, err
-		}
-		extKey, err := globalCacheKMS.DecryptKey(keyID, kmsKey, kms.Context{c.Bucket: path.Join(c.Bucket, c.Object)})
-		if err != nil {
-			return partETags, err
-		}
-		var objectKey crypto.ObjectKey
-		if err = objectKey.Unseal(extKey, sealedKey, crypto.S3.String(), c.Bucket, c.Object); err != nil {
-			return partETags, err
 		}
+		etags := make([]string, 0, len(c.PartETags))
 		for i := range c.PartETags {
-			etagStr := tryDecryptETag(objectKey[:], c.PartETags[i], false)
+			ETag, err := etag.Parse(c.PartETags[i])
-			// backend ETag was hex encoded before encrypting, so hex decode to get actual ETag
-			etag, err := hex.DecodeString(etagStr)
 			if err != nil {
-				return []string{}, err
+				return nil, err
 			}
-			partETags = append(partETags, string(etag))
+			ETag, err = etag.Decrypt(key[:], ETag)
+			if err != nil {
+				return nil, err
 			}
-		return partETags, nil
+			etags = append(etags, ETag.Format().String())
+		}
+		return etags, nil
+	case ok && (kind == crypto.S3KMS || kind == crypto.SSEC) && isCacheEncrypted(c.Meta):
+		etags := make([]string, 0, len(c.PartETags))
+		for i := range c.PartETags {
+			ETag, err := etag.Parse(c.PartETags[i])
+			if err != nil {
+				return nil, err
+			}
+			etags = append(etags, ETag.Format().String())
+		}
+		return etags, nil
 	default:
 		return c.PartETags, nil
 	}

cmd/disk-cache.go

@@ -31,8 +31,8 @@ import (
 	objectlock "github.com/minio/minio/internal/bucket/object/lock"
 	"github.com/minio/minio/internal/color"
 	"github.com/minio/minio/internal/config/cache"
+	"github.com/minio/minio/internal/disk"
 	"github.com/minio/minio/internal/hash"
-	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/minio/internal/sync/errgroup"
 	"github.com/minio/pkg/wildcard"
@@ -594,12 +594,23 @@ func newCache(config cache.Config) ([]*diskCache, bool, error) {
 	if err != nil {
 		return nil, false, err
 	}
+	var warningMsg string
 	for i, dir := range config.Drives {
 		// skip diskCache creation for cache drives missing a format.json
 		if formats[i] == nil {
 			caches = append(caches, nil)
 			continue
 		}
+		if !globalIsCICD && len(warningMsg) == 0 {
+			rootDsk, err := disk.IsRootDisk(dir, "/")
+			if err != nil {
+				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)
+			}
+		}
+
 		if err := checkAtimeSupport(dir); err != nil {
 			return nil, false, fmt.Errorf("Atime support required for disk caching, atime check failed with %w", err)
 		}
@@ -610,11 +621,14 @@ 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)))
+	}
 	return caches, migrating, nil
 }
 
 func (c *cacheObjects) migrateCacheFromV1toV2(ctx context.Context) {
-	logStartupMessage(color.Blue("Cache migration initiated ...."))
+	logger.Info(color.Blue("Cache migration initiated ...."))
 
 	g := errgroup.WithNErrs(len(c.cache))
 	for index, dc := range c.cache {
@@ -643,7 +657,7 @@ func (c *cacheObjects) migrateCacheFromV1toV2(ctx context.Context) {
 
 	// update migration status
 	c.migrating = false
-	logStartupMessage(color.Blue("Cache migration completed successfully."))
+	logger.Info(color.Blue("Cache migration completed successfully."))
 }
 
 // PutObject - caches the uploaded object for single Put operations
@@ -724,35 +738,30 @@ func (c *cacheObjects) PutObject(ctx context.Context, bucket, object string, r *
 	// Initialize pipe to stream data to cache
 	rPipe, wPipe := io.Pipe()
 	infoCh := make(chan ObjectInfo)
-	errorCh := make(chan error)
 	go func() {
+		defer close(infoCh)
 		info, err := putObjectFn(ctx, bucket, object, NewPutObjReader(hashReader), opts)
-		if err != nil {
-			close(infoCh)
 		pipeReader.CloseWithError(err)
 		rPipe.CloseWithError(err)
-			errorCh <- err
+		if err == nil {
-			return
-		}
-		close(errorCh)
 			infoCh <- info
+		}
 	}()
 
 	go func() {
 		_, err := dcache.put(lkctx.Context(), bucket, object, rPipe, r.Size(), nil, opts, false, false)
 		if err != nil {
-			rPipe.CloseWithError(err)
+			logger.LogIf(lkctx.Context(), err)
-			return
 		}
+		// We do not care about errors to cached backend.
+		rPipe.Close()
 	}()
 
 	mwriter := cacheMultiWriter(pipeWriter, wPipe)
 	_, err = io.Copy(mwriter, r)
 	pipeWriter.Close()
 	wPipe.Close()
-
 	if err != nil {
-		err = <-errorCh
 		return ObjectInfo{}, err
 	}
 	info := <-infoCh
@@ -788,8 +797,7 @@ func (c *cacheObjects) uploadObject(ctx context.Context, oi ObjectInfo) {
 		return
 	}
 	var opts ObjectOptions
-	opts.UserDefined = make(map[string]string)
+	opts.UserDefined = cloneMSS(oi.UserDefined)
-	opts.UserDefined[xhttp.ContentMD5] = oi.UserDefined["content-md5"]
 	objInfo, err := c.InnerPutObjectFn(ctx, oi.Bucket, oi.Name, NewPutObjReader(hashReader), opts)
 	wbCommitStatus := CommitComplete
 	size := objInfo.Size

cmd/encryption-v1.go

@@ -19,6 +19,7 @@ package cmd
 
 import (
 	"bufio"
+	"context"
 	"crypto/hmac"
 	"crypto/rand"
 	"crypto/sha256"
@@ -35,6 +36,7 @@ import (
 
 	"github.com/minio/kes"
 	"github.com/minio/minio/internal/crypto"
+	"github.com/minio/minio/internal/etag"
 	"github.com/minio/minio/internal/fips"
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/kms"
@@ -71,18 +73,132 @@ const (
 )
 
 // KMSKeyID returns in AWS compatible KMS KeyID() format.
-func (o ObjectInfo) KMSKeyID() string {
+func (o *ObjectInfo) KMSKeyID() string { return kmsKeyIDFromMetadata(o.UserDefined) }
-	if len(o.UserDefined) == 0 {
+
+// KMSKeyID returns in AWS compatible KMS KeyID() format.
+func (o *MultipartInfo) KMSKeyID() string { return kmsKeyIDFromMetadata(o.UserDefined) }
+
+// kmsKeyIDFromMetadata returns any AWS S3 KMS key ID in the
+// 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:"
+	if len(metadata) == 0 {
 		return ""
 	}
-	kmsID, ok := o.UserDefined[crypto.MetaKeyID]
+	kmsID, ok := metadata[crypto.MetaKeyID]
 	if !ok {
 		return ""
 	}
-	if strings.HasPrefix(kmsID, "arn:aws:kms:") {
+	if strings.HasPrefix(kmsID, ARNPrefix) {
 		return kmsID
 	}
-	return "arn:aws:kms:" + kmsID
+	return ARNPrefix + kmsID
+}
+
+// DecryptETags decryptes the ETag of all ObjectInfos using the KMS.
+//
+// It adjusts the size of all encrypted objects since encrypted
+// objects are slightly larger due to encryption overhead.
+// Further, it decrypts all single-part SSE-S3 encrypted objects
+// and formats ETags of SSE-C / SSE-KMS encrypted objects to
+// be AWS S3 compliant.
+//
+// DecryptETags uses a KMS bulk decryption API, if available, which
+// is more efficient than decrypting ETags sequentually.
+func DecryptETags(ctx context.Context, KMS kms.KMS, objects []ObjectInfo) error {
+	const BatchSize = 250 // We process the objects in batches - 250 is a reasonable default.
+	var (
+		metadata = make([]map[string]string, 0, BatchSize)
+		buckets  = make([]string, 0, BatchSize)
+		names    = make([]string, 0, BatchSize)
+	)
+	for len(objects) > 0 {
+		N := BatchSize
+		if len(objects) < BatchSize {
+			N = len(objects)
+		}
+		batch := objects[:N]
+
+		// We have to decrypt only ETags of SSE-S3 single-part
+		// objects.
+		// Therefore, we remember which objects (there index)
+		// in the current batch are single-part SSE-S3 objects.
+		metadata = metadata[:0:N]
+		buckets = buckets[:0:N]
+		names = names[:0:N]
+		SSES3SinglePartObjects := make(map[int]bool)
+		for i, object := range batch {
+			if kind, ok := crypto.IsEncrypted(object.UserDefined); ok && kind == crypto.S3 && !crypto.IsMultiPart(object.UserDefined) {
+				SSES3SinglePartObjects[i] = true
+
+				metadata = append(metadata, object.UserDefined)
+				buckets = append(buckets, object.Bucket)
+				names = append(names, object.Name)
+			}
+		}
+
+		// If there are no SSE-S3 single-part objects
+		// we can skip the decryption process. However,
+		// we still have to adjust the size and ETag
+		// of SSE-C and SSE-KMS objects.
+		if len(SSES3SinglePartObjects) == 0 {
+			for i := range batch {
+				size, err := batch[i].GetActualSize()
+				if err != nil {
+					return err
+				}
+				batch[i].Size = size
+
+				if _, ok := crypto.IsEncrypted(batch[i].UserDefined); ok {
+					ETag, err := etag.Parse(batch[i].ETag)
+					if err != nil {
+						return err
+					}
+					batch[i].ETag = ETag.Format().String()
+				}
+			}
+			objects = objects[N:]
+			continue
+		}
+
+		// There is at least one SSE-S3 single-part object.
+		// For all SSE-S3 single-part objects we have to
+		// fetch their decryption keys. We do this using
+		// a Bulk-Decryption API call, if available.
+		keys, err := crypto.S3.UnsealObjectKeys(ctx, KMS, metadata, buckets, names)
+		if err != nil {
+			return err
+		}
+
+		// Now, we have to decrypt the ETags of SSE-S3 single-part
+		// objects and adjust the size and ETags of all encrypted
+		// objects.
+		for i := range batch {
+			size, err := batch[i].GetActualSize()
+			if err != nil {
+				return err
+			}
+			batch[i].Size = size
+
+			if _, ok := crypto.IsEncrypted(batch[i].UserDefined); ok {
+				ETag, err := etag.Parse(batch[i].ETag)
+				if err != nil {
+					return err
+				}
+				if SSES3SinglePartObjects[i] && ETag.IsEncrypted() {
+					ETag, err = etag.Decrypt(keys[0][:], ETag)
+					if err != nil {
+						return err
+					}
+					keys = keys[1:]
+				}
+				batch[i].ETag = ETag.Format().String()
+			}
+		}
+		objects = objects[N:]
+	}
+	return nil
 }
 
 // isMultipart returns true if the current object is

cmd/endpoint.go

@@ -582,7 +582,7 @@ func CreateEndpoints(serverAddr string, foundLocal bool, args ...[]string) (Endp
 			return endpoints, setupType, config.ErrInvalidFSEndpoint(nil).Msg("use path style endpoint for FS setup")
 		}
 		endpoints = append(endpoints, endpoint)
-		setupType = FSSetupType
+		setupType = ErasureSDSetupType
 
 		// Check for cross device mounts if any.
 		if err = checkCrossDeviceMounts(endpoints); err != nil {

cmd/endpoint_test.go

@@ -231,10 +231,10 @@ func TestCreateEndpoints(t *testing.T) {
 	}{
 		{"localhost", [][]string{}, "", Endpoints{}, -1, fmt.Errorf("address localhost: missing port in address")},
 
-		// FS Setup
+		// Erasure Single Drive
 		{"localhost:9000", [][]string{{"http://localhost/d1"}}, "", Endpoints{}, -1, fmt.Errorf("use path style endpoint for FS setup")},
-		{":443", [][]string{{"/d1"}}, ":443", Endpoints{Endpoint{URL: &url.URL{Path: mustAbs("/d1")}, IsLocal: true}}, FSSetupType, nil},
+		{":443", [][]string{{"/d1"}}, ":443", Endpoints{Endpoint{URL: &url.URL{Path: mustAbs("/d1")}, IsLocal: true}}, ErasureSDSetupType, nil},
-		{"localhost:10000", [][]string{{"/d1"}}, "localhost:10000", Endpoints{Endpoint{URL: &url.URL{Path: mustAbs("/d1")}, IsLocal: true}}, FSSetupType, nil},
+		{"localhost:10000", [][]string{{"/d1"}}, "localhost:10000", Endpoints{Endpoint{URL: &url.URL{Path: mustAbs("/d1")}, IsLocal: true}}, ErasureSDSetupType, nil},
 		{"localhost:9000", [][]string{{"https://127.0.0.1:9000/d1", "https://localhost:9001/d1", "https://example.com/d1", "https://example.com/d2"}}, "", Endpoints{}, -1, fmt.Errorf("path '/d1' can not be served by different port on same address")},
 
 		// Erasure Setup with PathEndpointType

cmd/erasure-bucket.go

@@ -39,9 +39,11 @@ func (er erasureObjects) MakeBucketWithLocation(ctx context.Context, bucket stri
 	defer NSUpdated(bucket, slashSeparator)
 
 	// Verify if bucket is valid.
+	if !isMinioMetaBucketName(bucket) {
 		if err := s3utils.CheckValidBucketNameStrict(bucket); err != nil {
 			return BucketNameInvalid{Bucket: bucket}
 		}
+	}
 
 	storageDisks := er.getDisks()
 
@@ -53,6 +55,11 @@ func (er erasureObjects) MakeBucketWithLocation(ctx context.Context, bucket stri
 		g.Go(func() error {
 			if storageDisks[index] != nil {
 				if err := storageDisks[index].MakeVol(ctx, bucket); err != nil {
+					if opts.ForceCreate && errors.Is(err, errVolumeExists) {
+						// No need to return error when force create was
+						// requested.
+						return nil
+					}
 					if !errors.Is(err, errVolumeExists) {
 						logger.LogIf(ctx, err)
 					}

cmd/erasure-coding.go

@@ -41,7 +41,7 @@ type Erasure struct {
 // NewErasure creates a new ErasureStorage.
 func NewErasure(ctx context.Context, dataBlocks, parityBlocks int, blockSize int64) (e Erasure, err error) {
 	// Check the parameters for sanity now.
-	if dataBlocks <= 0 || parityBlocks <= 0 {
+	if dataBlocks <= 0 || parityBlocks < 0 {
 		return e, reedsolomon.ErrInvShardNum
 	}
 

cmd/erasure-healing-common.go

@@ -31,7 +31,7 @@ func commonTimeAndOccurence(times []time.Time, group time.Duration) (maxTime tim
 	groupNano := group.Nanoseconds()
 	// Ignore the uuid sentinel and count the rest.
 	for _, t := range times {
-		if t.Equal(timeSentinel) {
+		if t.Equal(timeSentinel) || t.IsZero() {
 			continue
 		}
 		nano := t.UnixNano()
@@ -83,7 +83,10 @@ func commonTime(modTimes []time.Time) (modTime time.Time) {
 }
 
 // Beginning of unix time is treated as sentinel value here.
-var timeSentinel = time.Unix(0, 0).UTC()
+var (
+	timeSentinel     = time.Unix(0, 0).UTC()
+	timeSentinel1970 = time.Unix(0, 1).UTC() // 1970 used for special cases when xlmeta.version == 0
+)
 
 // Boot modTimes up to disk count, setting the value to time sentinel.
 func bootModtimes(diskCount int) []time.Time {

cmd/erasure-healing.go

@@ -60,7 +60,8 @@ func (fi FileInfo) DataShardFixed() bool {
 // also heals the missing entries for bucket metadata files
 // `policy.json, notification.xml, listeners.json`.
 func (er erasureObjects) HealBucket(ctx context.Context, bucket string, opts madmin.HealOpts) (
-	result madmin.HealResultItem, err error) {
+	result madmin.HealResultItem, err error,
+) {
 	if !opts.DryRun {
 		defer NSUpdated(bucket, slashSeparator)
 	}
@@ -417,10 +418,19 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		return result, nil
 	}
 
+	if !latestMeta.XLV1 && !latestMeta.Deleted && 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)
+		logger.LogIf(ctx, err)
+		return er.defaultHealResult(latestMeta, storageDisks, storageEndpoints, errs,
+			bucket, object, versionID), err
+	}
+
 	cleanFileInfo := func(fi FileInfo) FileInfo {
-		// Returns a copy of the 'fi' with checksums and parts nil'ed.
+		// Returns a copy of the 'fi' with erasure index, checksums and inline data niled.
 		nfi := fi
-		if !fi.IsRemote() {
+		if !nfi.IsRemote() {
+			nfi.Data = nil
 			nfi.Erasure.Index = 0
 			nfi.Erasure.Checksums = nil
 		}
@@ -431,12 +441,25 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 	tmpID := mustGetUUID()
 	migrateDataDir := mustGetUUID()
 
+	// Reorder so that we have data disks first and parity disks next.
+	latestDisks := shuffleDisks(availableDisks, latestMeta.Erasure.Distribution)
+	outDatedDisks = shuffleDisks(outDatedDisks, latestMeta.Erasure.Distribution)
+	partsMetadata = shufflePartsMetadata(partsMetadata, latestMeta.Erasure.Distribution)
+
 	copyPartsMetadata := make([]FileInfo, len(partsMetadata))
+	for i := range latestDisks {
+		if latestDisks[i] == nil {
+			continue
+		}
+		copyPartsMetadata[i] = partsMetadata[i]
+	}
+
 	for i := range outDatedDisks {
 		if outDatedDisks[i] == nil {
 			continue
 		}
-		copyPartsMetadata[i] = partsMetadata[i]
+		// Make sure to write the FileInfo information
+		// that is expected to be in quorum.
 		partsMetadata[i] = cleanFileInfo(latestMeta)
 	}
 
@@ -455,12 +478,6 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 		inlineBuffers = make([]*bytes.Buffer, len(outDatedDisks))
 	}
 
-	// Reorder so that we have data disks first and parity disks next.
-	latestDisks := shuffleDisks(availableDisks, latestMeta.Erasure.Distribution)
-	outDatedDisks = shuffleDisks(outDatedDisks, latestMeta.Erasure.Distribution)
-	partsMetadata = shufflePartsMetadata(partsMetadata, latestMeta.Erasure.Distribution)
-	copyPartsMetadata = shufflePartsMetadata(copyPartsMetadata, latestMeta.Erasure.Distribution)
-
 	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
@@ -485,7 +502,7 @@ func (er erasureObjects) healObject(ctx context.Context, bucket string, object s
 				}
 				checksumInfo := copyPartsMetadata[i].Erasure.GetChecksumInfo(partNumber)
 				partPath := pathJoin(object, srcDataDir, fmt.Sprintf("part.%d", partNumber))
-				readers[i] = newBitrotReader(disk, partsMetadata[i].Data, bucket, partPath, tillOffset, checksumAlgo,
+				readers[i] = newBitrotReader(disk, copyPartsMetadata[i].Data, bucket, partPath, tillOffset, checksumAlgo,
 					checksumInfo.Hash, erasure.ShardSize())
 			}
 			writers := make([]io.Writer, len(outDatedDisks))
@@ -931,5 +948,12 @@ func (er erasureObjects) HealObject(ctx context.Context, bucket, object, version
 	}
 
 	// Heal the object.
-	return er.healObject(healCtx, bucket, object, versionID, opts)
+	hr, err = er.healObject(healCtx, bucket, object, versionID, opts)
+	if errors.Is(err, errFileCorrupt) && opts.ScanMode != madmin.HealDeepScan {
+		// Instead of returning an error when a bitrot error is detected
+		// during a normal heal scan, heal again with bitrot flag enabled.
+		opts.ScanMode = madmin.HealDeepScan
+		hr, err = er.healObject(healCtx, bucket, object, versionID, opts)
+	}
+	return hr, err
 }

cmd/erasure-healing_test.go

@@ -382,6 +382,13 @@ func TestHealingDanglingObject(t *testing.T) {
 		copy(disks, newDisks)
 		objLayer.(*erasureServerPools).serverPools[0].erasureDisksMu.Unlock()
 	}
+	getDisk := func(n int) StorageAPI {
+		objLayer.(*erasureServerPools).serverPools[0].erasureDisksMu.Lock()
+		disk := disks[n]
+		objLayer.(*erasureServerPools).serverPools[0].erasureDisksMu.Unlock()
+		return disk
+	}
+
 	// Remove 4 disks.
 	setDisks(nil, nil, nil, nil)
 
@@ -440,8 +447,8 @@ func TestHealingDanglingObject(t *testing.T) {
 	}
 
 	setDisks(orgDisks[:4]...)
-
+	disk := getDisk(0)
-	fileInfoPreHeal, err = disks[0].ReadVersion(context.Background(), bucket, object, "", false)
+	fileInfoPreHeal, err = disk.ReadVersion(context.Background(), bucket, object, "", false)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -458,7 +465,8 @@ func TestHealingDanglingObject(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	fileInfoPostHeal, err = disks[0].ReadVersion(context.Background(), bucket, object, "", false)
+	disk = getDisk(0)
+	fileInfoPostHeal, err = disk.ReadVersion(context.Background(), bucket, object, "", false)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -488,7 +496,8 @@ func TestHealingDanglingObject(t *testing.T) {
 
 	setDisks(orgDisks[:4]...)
 
-	fileInfoPreHeal, err = disks[0].ReadVersion(context.Background(), bucket, object, "", false)
+	disk = getDisk(0)
+	fileInfoPreHeal, err = disk.ReadVersion(context.Background(), bucket, object, "", false)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -505,7 +514,8 @@ func TestHealingDanglingObject(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	fileInfoPostHeal, err = disks[0].ReadVersion(context.Background(), bucket, object, "", false)
+	disk = getDisk(0)
+	fileInfoPostHeal, err = disk.ReadVersion(context.Background(), bucket, object, "", false)
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -748,8 +758,10 @@ func TestHealObjectCorruptedPools(t *testing.T) {
 			t.Fatalf("Failed to getLatestFileInfo - %v", err)
 		}
 
+		fi.DiskMTime = time.Time{}
+		nfi.DiskMTime = time.Time{}
 		if !reflect.DeepEqual(fi, nfi) {
-			t.Fatalf("FileInfo not equal after healing")
+			t.Fatalf("FileInfo not equal after healing: %v != %v", fi, nfi)
 		}
 
 		err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), false)
@@ -774,8 +786,10 @@ func TestHealObjectCorruptedPools(t *testing.T) {
 			t.Fatalf("Failed to getLatestFileInfo - %v", err)
 		}
 
+		fi.DiskMTime = time.Time{}
+		nfi.DiskMTime = time.Time{}
 		if !reflect.DeepEqual(fi, nfi) {
-			t.Fatalf("FileInfo not equal after healing")
+			t.Fatalf("FileInfo not equal after healing: %v != %v", fi, nfi)
 		}
 
 		// Test 4: checks if HealObject returns an error when xl.meta is not found
@@ -807,7 +821,144 @@ func TestHealObjectCorruptedPools(t *testing.T) {
 	}
 }
 
-func TestHealObjectCorrupted(t *testing.T) {
+func TestHealObjectCorruptedXLMeta(t *testing.T) {
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	resetGlobalHealState()
+	defer resetGlobalHealState()
+
+	nDisks := 16
+	fsDirs, err := getRandomDisks(nDisks)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	defer removeRoots(fsDirs)
+
+	// Everything is fine, should return nil
+	objLayer, _, err := initObjectLayer(ctx, mustGetPoolEndpoints(fsDirs...))
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	bucket := getRandomBucketName()
+	object := getRandomObjectName()
+	data := bytes.Repeat([]byte("a"), 5*1024*1024)
+	var opts ObjectOptions
+
+	err = objLayer.MakeBucketWithLocation(ctx, bucket, BucketOptions{})
+	if err != nil {
+		t.Fatalf("Failed to make a bucket - %v", err)
+	}
+
+	// Create an object with multiple parts uploaded in decreasing
+	// part number.
+	uploadID, err := objLayer.NewMultipartUpload(ctx, bucket, object, opts)
+	if err != nil {
+		t.Fatalf("Failed to create a multipart upload - %v", err)
+	}
+
+	var uploadedParts []CompletePart
+	for _, partID := range []int{2, 1} {
+		pInfo, err1 := objLayer.PutObjectPart(ctx, bucket, object, uploadID, partID, mustGetPutObjReader(t, bytes.NewReader(data), int64(len(data)), "", ""), opts)
+		if err1 != nil {
+			t.Fatalf("Failed to upload a part - %v", err1)
+		}
+		uploadedParts = append(uploadedParts, CompletePart{
+			PartNumber: pInfo.PartNumber,
+			ETag:       pInfo.ETag,
+		})
+	}
+
+	_, err = objLayer.CompleteMultipartUpload(ctx, bucket, object, uploadID, uploadedParts, ObjectOptions{})
+	if err != nil {
+		t.Fatalf("Failed to complete multipart upload - %v", err)
+	}
+
+	z := objLayer.(*erasureServerPools)
+	er := z.serverPools[0].sets[0]
+	erasureDisks := er.getDisks()
+	firstDisk := erasureDisks[0]
+
+	// Test 1: Remove the object backend files from the first disk.
+	fileInfos, errs := readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	fi, err := getLatestFileInfo(ctx, fileInfos, errs)
+	if err != nil {
+		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+	}
+
+	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
+	if err != nil {
+		t.Fatalf("Failed to delete a file - %v", err)
+	}
+
+	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{ScanMode: madmin.HealNormalScan})
+	if err != nil {
+		t.Fatalf("Failed to heal object - %v", err)
+	}
+
+	if _, err = firstDisk.StatInfoFile(context.Background(), bucket, object+"/"+xlStorageFormatFile, false); err != nil {
+		t.Errorf("Expected xl.meta file to be present but stat failed - %v", err)
+	}
+
+	fileInfos, errs = readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	nfi1, err := getLatestFileInfo(ctx, fileInfos, errs)
+	if err != nil {
+		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+	}
+
+	fi.DiskMTime = time.Time{}
+	nfi1.DiskMTime = time.Time{}
+	if !reflect.DeepEqual(fi, nfi1) {
+		t.Fatalf("FileInfo not equal after healing")
+	}
+
+	// Test 2: Test with a corrupted xl.meta
+	err = firstDisk.WriteAll(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), []byte("abcd"))
+	if err != nil {
+		t.Errorf("Failure during creating part.1 - %v", err)
+	}
+
+	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{ScanMode: madmin.HealNormalScan})
+	if err != nil {
+		t.Errorf("Expected nil but received %v", err)
+	}
+
+	fileInfos, errs = readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	nfi2, err := getLatestFileInfo(ctx, fileInfos, errs)
+	if err != nil {
+		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+	}
+
+	fi.DiskMTime = time.Time{}
+	nfi2.DiskMTime = time.Time{}
+	if !reflect.DeepEqual(fi, nfi2) {
+		t.Fatalf("FileInfo not equal after healing")
+	}
+
+	// Test 3: checks if HealObject returns an error when xl.meta is not found
+	// in more than read quorum number of disks, to create a corrupted situation.
+	for i := 0; i <= nfi2.Erasure.DataBlocks; i++ {
+		erasureDisks[i].Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
+	}
+
+	// Try healing now, expect to receive errFileNotFound.
+	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{DryRun: false, Remove: true, ScanMode: madmin.HealDeepScan})
+	if err != nil {
+		if _, ok := err.(ObjectNotFound); !ok {
+			t.Errorf("Expect %v but received %v", ObjectNotFound{Bucket: bucket, Object: object}, err)
+		}
+	}
+
+	// since majority of xl.meta's are not available, object should be successfully deleted.
+	_, err = objLayer.GetObjectInfo(ctx, bucket, object, ObjectOptions{})
+	if _, ok := err.(ObjectNotFound); !ok {
+		t.Errorf("Expect %v but received %v", ObjectNotFound{Bucket: bucket, Object: object}, err)
+	}
+}
+
+func TestHealObjectCorruptedParts(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
@@ -867,7 +1018,26 @@ func TestHealObjectCorrupted(t *testing.T) {
 	er := z.serverPools[0].sets[0]
 	erasureDisks := er.getDisks()
 	firstDisk := erasureDisks[0]
-	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
+	secondDisk := erasureDisks[1]
+
+	fileInfos, errs := readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	fi, err := getLatestFileInfo(ctx, fileInfos, errs)
+	if err != nil {
+		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+	}
+
+	part1Disk1Origin, err := firstDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
+	if err != nil {
+		t.Fatalf("Failed to read a file - %v", err)
+	}
+
+	part1Disk2Origin, err := secondDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
+	if err != nil {
+		t.Fatalf("Failed to read a file - %v", err)
+	}
+
+	// Test 1, remove part.1
+	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), false)
 	if err != nil {
 		t.Fatalf("Failed to delete a file - %v", err)
 	}
@@ -877,85 +1047,67 @@ func TestHealObjectCorrupted(t *testing.T) {
 		t.Fatalf("Failed to heal object - %v", err)
 	}
 
-	fileInfos, errs := readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	part1Replaced, err := firstDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
-	fi, err := getLatestFileInfo(ctx, fileInfos, errs)
 	if err != nil {
-		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+		t.Fatalf("Failed to read a file - %v", err)
 	}
 
-	if _, err = firstDisk.StatInfoFile(context.Background(), bucket, object+"/"+xlStorageFormatFile, false); err != nil {
+	if !reflect.DeepEqual(part1Disk1Origin, part1Replaced) {
-		t.Errorf("Expected xl.meta file to be present but stat failed - %v", err)
+		t.Fatalf("part.1 not healed correctly")
 	}
 
-	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), false)
+	// Test 2, Corrupt part.1
+	err = firstDisk.WriteAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), []byte("foobytes"))
 	if err != nil {
-		t.Errorf("Failure during deleting part.1 - %v", err)
+		t.Fatalf("Failed to write a file - %v", err)
 	}
 
-	err = firstDisk.WriteAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), []byte{})
+	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{ScanMode: madmin.HealNormalScan})
 	if err != nil {
-		t.Errorf("Failure during creating part.1 - %v", err)
+		t.Fatalf("Failed to heal object - %v", err)
 	}
 
-	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{DryRun: false, Remove: true, ScanMode: madmin.HealDeepScan})
+	part1Replaced, err = firstDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
 	if err != nil {
-		t.Errorf("Expected nil but received %v", err)
+		t.Fatalf("Failed to read a file - %v", err)
 	}
 
-	fileInfos, errs = readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	if !reflect.DeepEqual(part1Disk1Origin, part1Replaced) {
-	nfi, err := getLatestFileInfo(ctx, fileInfos, errs)
+		t.Fatalf("part.1 not healed correctly")
+	}
+
+	// Test 3, Corrupt one part and remove data in another disk
+	err = firstDisk.WriteAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), []byte("foobytes"))
 	if err != nil {
-		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+		t.Fatalf("Failed to write a file - %v", err)
 	}
 
-	if !reflect.DeepEqual(fi, nfi) {
+	err = secondDisk.Delete(context.Background(), bucket, object, true)
-		t.Fatalf("FileInfo not equal after healing")
-	}
-
-	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), false)
 	if err != nil {
-		t.Errorf("Failure during deleting part.1 - %v", err)
+		t.Fatalf("Failed to delete a file - %v", err)
 	}
 
-	bdata := bytes.Repeat([]byte("b"), int(nfi.Size))
+	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{ScanMode: madmin.HealNormalScan})
-	err = firstDisk.WriteAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"), bdata)
 	if err != nil {
-		t.Errorf("Failure during creating part.1 - %v", err)
+		t.Fatalf("Failed to heal object - %v", err)
 	}
 
-	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{DryRun: false, Remove: true, ScanMode: madmin.HealDeepScan})
+	partReconstructed, err := firstDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
 	if err != nil {
-		t.Errorf("Expected nil but received %v", err)
+		t.Fatalf("Failed to read a file - %v", err)
 	}
 
-	fileInfos, errs = readAllFileInfo(ctx, erasureDisks, bucket, object, "", false)
+	if !reflect.DeepEqual(part1Disk1Origin, partReconstructed) {
-	nfi, err = getLatestFileInfo(ctx, fileInfos, errs)
+		t.Fatalf("part.1 not healed correctly")
+	}
+
+	partReconstructed, err = secondDisk.ReadAll(context.Background(), bucket, pathJoin(object, fi.DataDir, "part.1"))
 	if err != nil {
-		t.Fatalf("Failed to getLatestFileInfo - %v", err)
+		t.Fatalf("Failed to read a file - %v", err)
 	}
 
-	if !reflect.DeepEqual(fi, nfi) {
+	if !reflect.DeepEqual(part1Disk2Origin, partReconstructed) {
-		t.Fatalf("FileInfo not equal after healing")
+		t.Fatalf("part.1 not healed correctly")
-	}
-
-	// Test 4: checks if HealObject returns an error when xl.meta is not found
-	// in more than read quorum number of disks, to create a corrupted situation.
-	for i := 0; i <= nfi.Erasure.DataBlocks; i++ {
-		erasureDisks[i].Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
-	}
-
-	// Try healing now, expect to receive errFileNotFound.
-	_, err = objLayer.HealObject(ctx, bucket, object, "", madmin.HealOpts{DryRun: false, Remove: true, ScanMode: madmin.HealDeepScan})
-	if err != nil {
-		if _, ok := err.(ObjectNotFound); !ok {
-			t.Errorf("Expect %v but received %v", ObjectNotFound{Bucket: bucket, Object: object}, err)
-		}
-	}
-
-	// since majority of xl.meta's are not available, object should be successfully deleted.
-	_, err = objLayer.GetObjectInfo(ctx, bucket, object, ObjectOptions{})
-	if _, ok := err.(ObjectNotFound); !ok {
-		t.Errorf("Expect %v but received %v", ObjectNotFound{Bucket: bucket, Object: object}, err)
 	}
 }
 
@@ -1017,7 +1169,8 @@ func TestHealObjectErasure(t *testing.T) {
 		t.Fatalf("Failed to complete multipart upload - %v", err)
 	}
 
-	err = firstDisk.Delete(context.Background(), bucket, pathJoin(object, xlStorageFormatFile), false)
+	// Delete the whole object folder
+	err = firstDisk.Delete(context.Background(), bucket, object, true)
 	if err != nil {
 		t.Fatalf("Failed to delete a file - %v", err)
 	}

cmd/erasure-metadata-utils.go

@@ -137,7 +137,15 @@ func readAllFileInfo(ctx context.Context, disks []StorageAPI, bucket, object, ve
 				return errDiskNotFound
 			}
 			metadataArray[index], err = disks[index].ReadVersion(ctx, bucket, object, versionID, readData)
-			if err != nil {
+			return err
+		}, index)
+	}
+
+	errs := g.Wait()
+	for index, err := range errs {
+		if err == nil {
+			continue
+		}
 		if !IsErr(err, []error{
 			errFileNotFound,
 			errVolumeNotFound,
@@ -149,12 +157,9 @@ func readAllFileInfo(ctx context.Context, disks []StorageAPI, bucket, object, ve
 				disks[index].String())
 		}
 	}
-			return err
-		}, index)
-	}
 
 	// Return all the metadata.
-	return metadataArray, g.Wait()
+	return metadataArray, errs
 }
 
 // shuffleDisksAndPartsMetadataByIndex this function should be always used by GetObjectNInfo()
@@ -178,10 +183,6 @@ func shuffleDisksAndPartsMetadataByIndex(disks []StorageAPI, metaArr []FileInfo,
 			inconsistent++
 			continue
 		}
-		if len(fi.Data) != len(meta.Data) {
-			inconsistent++
-			continue
-		}
 		if meta.XLV1 != fi.XLV1 {
 			inconsistent++
 			continue
@@ -229,12 +230,6 @@ func shuffleDisksAndPartsMetadata(disks []StorageAPI, partsMetadata []FileInfo,
 			// if object was ever written previously.
 			continue
 		}
-		if !init && len(fi.Data) != len(partsMetadata[index].Data) {
-			// Check for length of data parts only when
-			// fi.ModTime is not empty - ModTime is always set,
-			// if object was ever written previously.
-			continue
-		}
 		if !init && fi.XLV1 != partsMetadata[index].XLV1 {
 			continue
 		}