metrics: fix typo in namespace for proxy tagging metric (#19039)

Relevant PR introducing this metric: #18957

.codespellrc

@@ -0,0 +1,8 @@
+
+[codespell]
+# certs_test.go - has lots of ceritificates.
+skip = go.mod,go.sum,*.txt,LICENSE,*.zip,.git,*.pdf,*.svg,.codespellrc,CREDITS,certs_test.go
+check-hidden = true
+ignore-regex = \b(newfolder/afile|filterIn|HelpES)\b
+ignore-words-list = inout,bui,to,bu,te,ot,toi,ist,parms,flate
+

.github/PULL_REQUEST_TEMPLATE.md

@@ -1,6 +1,6 @@
 ## Community Contribution License
 All community contributions in this pull request are licensed to the project maintainers
-under the terms of the [Apache 2 license] (https://www.apache.org/licenses/LICENSE-2.0). 
+under the terms of the [Apache 2 license](https://www.apache.org/licenses/LICENSE-2.0). 
 By creating this pull request I represent that I have the right to license the 
 contributions to the project maintainers under the Apache 2 license.
 

.github/workflows/codespell.yml

@@ -0,0 +1,20 @@
+---
+name: Codespell
+
+on:
+  pull_request:
+    branches: [master]
+
+permissions:
+  contents: read
+
+jobs:
+  codespell:
+    name: Check for spelling errors
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+      - name: Codespell
+        uses: codespell-project/actions-codespell@v2

.github/workflows/go-cross.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/go-fips.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/go-healing.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/go-lint.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/go.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/helm-lint.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/iam-integrations.yaml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/issues.yaml

@@ -0,0 +1,18 @@
+# @format
+
+name: Issue Workflow
+
+on:
+  issues:
+    types:
+      - opened
+
+jobs:
+  add-to-project:
+    name: Add issue to project
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/add-to-project@v0.5.0
+        with:
+          project-url: https://github.com/orgs/miniohq/projects/2
+          github-token: ${{ secrets.BOT_PAT }}

.github/workflows/mint.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
       - master
+      - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.
@@ -37,7 +38,11 @@ jobs:
 
       - name: build-minio
         run: |
-          TAG="minio/minio:${{ steps.vars.outputs.sha_short }}" make docker
+          TAG="quay.io/minio/minio:${{ steps.vars.outputs.sha_short }}" make docker
+
+      - name: multipart uploads test
+        run: |
+          ${GITHUB_WORKSPACE}/.github/workflows/multipart/migrate.sh "${{ steps.vars.outputs.sha_short }}"
 
       - name: compress and encrypt
         run: |
@@ -59,7 +64,14 @@ jobs:
              docker-compose -f ${GITHUB_WORKSPACE}/.github/workflows/mint/minio-${mode}.yaml down || true
              docker-compose -f ${GITHUB_WORKSPACE}/.github/workflows/mint/minio-${mode}.yaml rm || true
           done
-          docker rmi -f minio/minio:${{ steps.vars.outputs.sha_short }}
+
+          docker-compose -f ${GITHUB_WORKSPACE}/.github/workflows/multipart/docker-compose-site1.yaml rm -s -f || true
+          docker-compose -f ${GITHUB_WORKSPACE}/.github/workflows/multipart/docker-compose-site2.yaml rm -s -f || true
+          for volume in $(docker volume ls -q | grep minio); do
+            docker volume rm ${volume} || true
+          done
+
+          docker rmi -f quay.io/minio/minio:${{ steps.vars.outputs.sha_short }}
           docker system prune -f || true
           docker volume prune -f || true
           docker volume rm $(docker volume ls -q -f dangling=true) || true

.github/workflows/mint/minio-compress-encrypt.yaml

@@ -2,7 +2,7 @@ version: '3.7'
 
 # Settings and configurations that are common for all containers
 x-minio-common: &minio-common
-  image: minio/minio:${JOB_NAME}
+  image: quay.io/minio/minio:${JOB_NAME}
   command: server --console-address ":9001" http://minio{1...4}/cdata{1...2}
   expose:
     - "9000"
@@ -16,10 +16,10 @@ x-minio-common: &minio-common
     MINIO_COMPRESSION_ALLOW_ENCRYPTION: "on"
     MINIO_KMS_SECRET_KEY: "my-minio-key:OSMM+vkKUTCvQs9YL/CVMIMt43HFhkUpqJxTmGl6rYw="
   healthcheck:
-    test: ["CMD", "curl", "-f", "http://localhost:9000/minio/health/live"]
-    interval: 30s
-    timeout: 20s
-    retries: 3
+    test: ["CMD", "mc", "ready", "local"]
+    interval: 5s
+    timeout: 5s
+    retries: 5
 
 # starts 4 docker containers running minio server instances.
 # using nginx reverse proxy, load balancing, you can access

.github/workflows/mint/minio-erasure.yaml

@@ -2,7 +2,7 @@ version: '3.7'
 
 # Settings and configurations that are common for all containers
 x-minio-common: &minio-common
-  image: minio/minio:${JOB_NAME}
+  image: quay.io/minio/minio:${JOB_NAME}
   command: server --console-address ":9001" edata{1...4}
   expose:
     - "9000"
@@ -13,10 +13,10 @@ x-minio-common: &minio-common
     MINIO_ROOT_PASSWORD: "minio123"
     MINIO_KMS_SECRET_KEY: "my-minio-key:OSMM+vkKUTCvQs9YL/CVMIMt43HFhkUpqJxTmGl6rYw="
   healthcheck:
-    test: ["CMD", "curl", "-f", "http://localhost:9000/minio/health/live"]
-    interval: 30s
-    timeout: 20s
-    retries: 3
+    test: ["CMD", "mc", "ready", "local"]
+    interval: 5s
+    timeout: 5s
+    retries: 5
 
 # starts 4 docker containers running minio server instances.
 # using nginx reverse proxy, load balancing, you can access

.github/workflows/mint/minio-pools.yaml

@@ -2,7 +2,7 @@ version: '3.7'
 
 # Settings and configurations that are common for all containers
 x-minio-common: &minio-common
-  image: minio/minio:${JOB_NAME}
+  image: quay.io/minio/minio:${JOB_NAME}
   command: server --console-address ":9001" http://minio{1...4}/pdata{1...2} http://minio{5...8}/pdata{1...2}
   expose:
     - "9000"
@@ -13,10 +13,10 @@ x-minio-common: &minio-common
     MINIO_ROOT_PASSWORD: "minio123"
     MINIO_KMS_SECRET_KEY: "my-minio-key:OSMM+vkKUTCvQs9YL/CVMIMt43HFhkUpqJxTmGl6rYw="
   healthcheck:
-    test: ["CMD", "curl", "-f", "http://localhost:9000/minio/health/live"]
-    interval: 30s
-    timeout: 20s
-    retries: 3
+    test: ["CMD", "mc", "ready", "local"]
+    interval: 5s
+    timeout: 5s
+    retries: 5
 
 # starts 4 docker containers running minio server instances.
 # using nginx reverse proxy, load balancing, you can access

.github/workflows/multipart/docker-compose-site1.yaml

@@ -0,0 +1,66 @@
+version: '3.7'
+
+# Settings and configurations that are common for all containers
+x-minio-common: &minio-common
+  image: quay.io/minio/minio:${RELEASE}
+  command: server http://site1-minio{1...4}/data{1...2}
+  environment:
+    - MINIO_PROMETHEUS_AUTH_TYPE=public
+    - CI=true
+
+# starts 4 docker containers running minio server instances.
+# using nginx reverse proxy, load balancing, you can access
+# it through port 9000.
+services:
+  site1-minio1:
+    <<: *minio-common
+    hostname: site1-minio1
+    volumes:
+      - site1-data1-1:/data1
+      - site1-data1-2:/data2
+
+  site1-minio2:
+    <<: *minio-common
+    hostname: site1-minio2
+    volumes:
+      - site1-data2-1:/data1
+      - site1-data2-2:/data2
+
+  site1-minio3:
+    <<: *minio-common
+    hostname: site1-minio3
+    volumes:
+      - site1-data3-1:/data1
+      - site1-data3-2:/data2
+
+  site1-minio4:
+    <<: *minio-common
+    hostname: site1-minio4
+    volumes:
+      - site1-data4-1:/data1
+      - site1-data4-2:/data2
+
+  site1-nginx:
+    image: nginx:1.19.2-alpine
+    hostname: site1-nginx
+    volumes:
+      - ./nginx-site1.conf:/etc/nginx/nginx.conf:ro
+    ports:
+      - "9001:9001"
+    depends_on:
+      - site1-minio1
+      - site1-minio2
+      - site1-minio3
+      - site1-minio4
+
+## By default this config uses default local driver,
+## For custom volumes replace with volume driver configuration.
+volumes:
+  site1-data1-1:
+  site1-data1-2:
+  site1-data2-1:
+  site1-data2-2:
+  site1-data3-1:
+  site1-data3-2:
+  site1-data4-1:
+  site1-data4-2:

.github/workflows/multipart/docker-compose-site2.yaml

@@ -0,0 +1,66 @@
+version: '3.7'
+
+# Settings and configurations that are common for all containers
+x-minio-common: &minio-common
+  image: quay.io/minio/minio:${RELEASE}
+  command: server http://site2-minio{1...4}/data{1...2}
+  environment:
+    - MINIO_PROMETHEUS_AUTH_TYPE=public
+    - CI=true
+
+# starts 4 docker containers running minio server instances.
+# using nginx reverse proxy, load balancing, you can access
+# it through port 9000.
+services:
+  site2-minio1:
+    <<: *minio-common
+    hostname: site2-minio1
+    volumes:
+      - site2-data1-1:/data1
+      - site2-data1-2:/data2
+
+  site2-minio2:
+    <<: *minio-common
+    hostname: site2-minio2
+    volumes:
+      - site2-data2-1:/data1
+      - site2-data2-2:/data2
+
+  site2-minio3:
+    <<: *minio-common
+    hostname: site2-minio3
+    volumes:
+      - site2-data3-1:/data1
+      - site2-data3-2:/data2
+
+  site2-minio4:
+    <<: *minio-common
+    hostname: site2-minio4
+    volumes:
+      - site2-data4-1:/data1
+      - site2-data4-2:/data2
+
+  site2-nginx:
+    image: nginx:1.19.2-alpine
+    hostname: site2-nginx
+    volumes:
+      - ./nginx-site2.conf:/etc/nginx/nginx.conf:ro
+    ports:
+      - "9002:9002"
+    depends_on:
+      - site2-minio1
+      - site2-minio2
+      - site2-minio3
+      - site2-minio4
+
+## By default this config uses default local driver,
+## For custom volumes replace with volume driver configuration.
+volumes:
+  site2-data1-1:
+  site2-data1-2:
+  site2-data2-1:
+  site2-data2-2:
+  site2-data3-1:
+  site2-data3-2:
+  site2-data4-1:
+  site2-data4-2:

.github/workflows/multipart/migrate.sh

@@ -0,0 +1,115 @@
+#!/bin/bash
+
+set -x
+
+## change working directory
+cd .github/workflows/multipart/
+
+function cleanup() {
+	docker-compose -f docker-compose-site1.yaml rm -s -f || true
+	docker-compose -f docker-compose-site2.yaml rm -s -f || true
+	for volume in $(docker volume ls -q | grep minio); do
+		docker volume rm ${volume} || true
+	done
+
+	docker system prune -f || true
+	docker volume prune -f || true
+	docker volume rm $(docker volume ls -q -f dangling=true) || true
+}
+
+cleanup
+
+if [ ! -f ./mc ]; then
+	wget --quiet -O mc https://dl.minio.io/client/mc/release/linux-amd64/mc &&
+		chmod +x mc
+fi
+
+(
+	cd /tmp
+	go install github.com/minio/minio/docs/debugging/s3-check-md5@master
+)
+
+export RELEASE=RELEASE.2023-08-29T23-07-35Z
+
+docker-compose -f docker-compose-site1.yaml up -d
+docker-compose -f docker-compose-site2.yaml up -d
+
+sleep 30s
+
+./mc alias set site1 http://site1-nginx:9001 minioadmin minioadmin --api s3v4
+./mc alias set site2 http://site2-nginx:9002 minioadmin minioadmin --api s3v4
+
+./mc ready site1/
+./mc ready site2/
+
+./mc admin replicate add site1 site2
+./mc mb site1/testbucket/
+./mc cp -r --quiet /usr/bin site1/testbucket/
+
+sleep 5
+
+s3-check-md5 -h
+
+failed_count_site1=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site1-nginx:9001 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+failed_count_site2=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site2-nginx:9002 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+
+if [ $failed_count_site1 -ne 0 ]; then
+	echo "failed with multipart on site1 uploads"
+	exit 1
+fi
+
+if [ $failed_count_site2 -ne 0 ]; then
+	echo "failed with multipart on site2 uploads"
+	exit 1
+fi
+
+./mc cp -r --quiet /usr/bin site1/testbucket/
+
+sleep 5
+
+failed_count_site1=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site1-nginx:9001 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+failed_count_site2=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site2-nginx:9002 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+
+## we do not need to fail here, since we are going to test
+## upgrading to master, healing and being able to recover
+## the last version.
+if [ $failed_count_site1 -ne 0 ]; then
+	echo "failed with multipart on site1 uploads ${failed_count_site1}"
+fi
+
+if [ $failed_count_site2 -ne 0 ]; then
+	echo "failed with multipart on site2 uploads ${failed_count_site2}"
+fi
+
+export RELEASE=${1}
+
+docker-compose -f docker-compose-site1.yaml up -d
+docker-compose -f docker-compose-site2.yaml up -d
+
+./mc ready site1/
+./mc ready site2/
+
+for i in $(seq 1 10); do
+	# mc admin heal -r --remove when used against a LB endpoint
+	# behaves flaky, let this run 10 times before giving up
+	./mc admin heal -r --remove --json site1/ 2>&1 >/dev/null
+	./mc admin heal -r --remove --json site2/ 2>&1 >/dev/null
+done
+
+failed_count_site1=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site1-nginx:9001 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+failed_count_site2=$(s3-check-md5 -versions -access-key minioadmin -secret-key minioadmin -endpoint http://site2-nginx:9002 -bucket testbucket 2>&1 | grep FAILED | wc -l)
+
+if [ $failed_count_site1 -ne 0 ]; then
+	echo "failed with multipart on site1 uploads"
+	exit 1
+fi
+
+if [ $failed_count_site2 -ne 0 ]; then
+	echo "failed with multipart on site2 uploads"
+	exit 1
+fi
+
+cleanup
+
+## change working directory
+cd ../../../

.github/workflows/multipart/nginx-site1.conf

@@ -0,0 +1,61 @@
+user  nginx;
+worker_processes  auto;
+
+error_log  /var/log/nginx/error.log warn;
+pid        /var/run/nginx.pid;
+
+events {
+    worker_connections  4096;
+}
+
+http {
+    include       /etc/nginx/mime.types;
+    default_type  application/octet-stream;
+
+    log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
+                      '$status $body_bytes_sent "$http_referer" '
+                      '"$http_user_agent" "$http_x_forwarded_for"';
+
+    access_log  /var/log/nginx/access.log  main;
+    sendfile        on;
+    keepalive_timeout  65;
+
+    # include /etc/nginx/conf.d/*.conf;
+
+    upstream minio {
+        server site1-minio1:9000;
+        server site1-minio2:9000;
+        server site1-minio3:9000;
+        server site1-minio4:9000;
+    }
+
+    server {
+        listen       9001;
+        listen  [::]:9001;
+        server_name  localhost;
+
+        # To allow special characters in headers
+        ignore_invalid_headers off;
+        # Allow any size file to be uploaded.
+        # Set to a value such as 1000m; to restrict file size to a specific value
+        client_max_body_size 0;
+        # To disable buffering
+        proxy_buffering off;
+        proxy_request_buffering off;
+
+        location / {
+            proxy_set_header Host $http_host;
+            proxy_set_header X-Real-IP $remote_addr;
+            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+            proxy_set_header X-Forwarded-Proto $scheme;
+
+            proxy_connect_timeout 300;
+            # Default is HTTP/1, keepalive is only enabled in HTTP/1.1
+            proxy_http_version 1.1;
+            proxy_set_header Connection "";
+            chunked_transfer_encoding off;
+
+            proxy_pass http://minio;
+        }
+    }
+}

.github/workflows/multipart/nginx-site2.conf

@@ -0,0 +1,61 @@
+user  nginx;
+worker_processes  auto;
+
+error_log  /var/log/nginx/error.log warn;
+pid        /var/run/nginx.pid;
+
+events {
+    worker_connections  4096;
+}
+
+http {
+    include       /etc/nginx/mime.types;
+    default_type  application/octet-stream;
+
+    log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
+                      '$status $body_bytes_sent "$http_referer" '
+                      '"$http_user_agent" "$http_x_forwarded_for"';
+
+    access_log  /var/log/nginx/access.log  main;
+    sendfile        on;
+    keepalive_timeout  65;
+
+    # include /etc/nginx/conf.d/*.conf;
+
+    upstream minio {
+        server site2-minio1:9000;
+        server site2-minio2:9000;
+        server site2-minio3:9000;
+        server site2-minio4:9000;
+    }
+
+    server {
+        listen       9002;
+        listen  [::]:9002;
+        server_name  localhost;
+
+        # To allow special characters in headers
+        ignore_invalid_headers off;
+        # Allow any size file to be uploaded.
+        # Set to a value such as 1000m; to restrict file size to a specific value
+        client_max_body_size 0;
+        # To disable buffering
+        proxy_buffering off;
+        proxy_request_buffering off;
+
+        location / {
+            proxy_set_header Host $http_host;
+            proxy_set_header X-Real-IP $remote_addr;
+            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+            proxy_set_header X-Forwarded-Proto $scheme;
+
+            proxy_connect_timeout 300;
+            # Default is HTTP/1, keepalive is only enabled in HTTP/1.1
+            proxy_http_version 1.1;
+            proxy_set_header Connection "";
+            chunked_transfer_encoding off;
+
+            proxy_pass http://minio;
+        }
+    }
+}

.github/workflows/replication.yaml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.
@@ -35,6 +36,12 @@ jobs:
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make test-decom
 
+      - name: Test Config File
+        run: |
+          sudo sysctl net.ipv6.conf.all.disable_ipv6=0
+          sudo sysctl net.ipv6.conf.default.disable_ipv6=0
+          make test-configfile
+
       - name: Test Replication
         run: |
           sudo sysctl net.ipv6.conf.all.disable_ipv6=0
@@ -47,3 +54,8 @@ jobs:
           sudo sysctl net.ipv6.conf.default.disable_ipv6=0
           make test-site-replication-minio
 
+      - name: Test Versioning
+        run: |
+          sudo sysctl net.ipv6.conf.all.disable_ipv6=0
+          sudo sysctl net.ipv6.conf.default.disable_ipv6=0
+          make test-versioning

.github/workflows/root-disable.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/run-mint.sh

@@ -16,7 +16,7 @@ docker volume rm $(docker volume ls -f dangling=true) || true
 cd .github/workflows/mint
 
 docker-compose -f minio-${MODE}.yaml up -d
-sleep 5m
+sleep 30s
 
 docker system prune -f || true
 docker volume prune -f || true

.github/workflows/shfmt.yml

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 permissions:
   contents: read

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

@@ -4,6 +4,7 @@ on:
   pull_request:
     branches:
     - master
+    - next
 
 # This ensures that previous jobs for the PR are canceled when the PR is
 # updated.

.github/workflows/vulncheck.yml

@@ -3,6 +3,7 @@ on:
   pull_request:
     branches:
       - master
+
   push:
     branches:
       - master
@@ -20,7 +21,7 @@ jobs:
       - name: Set up Go
         uses: actions/setup-go@v3
         with:
-          go-version: 1.21.1
+          go-version: 1.21.5
           check-latest: true
       - name: Get official govulncheck
         run: go install golang.org/x/vuln/cmd/govulncheck@latest

.gitignore

@@ -42,3 +42,5 @@ docs/debugging/healing-bin/healing-bin
 docs/debugging/inspect/inspect
 docs/debugging/pprofgoparser/pprofgoparser
 docs/debugging/reorder-disks/reorder-disks
+docs/debugging/populate-hard-links/populate-hardlinks
+docs/debugging/xattr/xattr

.golangci.yml

@@ -13,7 +13,6 @@ linters:
   enable:
     - durationcheck
     - gocritic
-    - gofmt
     - gofumpt
     - goimports
     - gomodguard
@@ -30,5 +29,8 @@ linters:
 issues:
   exclude-use-default: false
   exclude:
+    - "empty-block:"
+    - "unused-parameter:"
+    - "dot-imports:"
     - should have a package comment
     - error strings should not be capitalized or end with punctuation or a newline

Dockerfile

@@ -1,8 +1,6 @@
 FROM minio/minio:latest
 
-ENV PATH=/opt/bin:$PATH
-
-COPY ./minio /opt/bin/minio
+COPY ./minio /usr/bin/minio
 COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 
 ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]

Dockerfile.hotfix

@@ -1,4 +1,31 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.8
+FROM golang:1.21-alpine as build
+
+ARG TARGETARCH
+ARG RELEASE
+
+ENV GOPATH /go
+ENV CGO_ENABLED 0
+
+# Install curl and minisign
+RUN apk add -U --no-cache ca-certificates && \
+    apk add -U --no-cache curl && \
+    go install aead.dev/minisign/cmd/minisign@v0.2.1
+
+# Download minio binary and signature file
+RUN curl -s -q https://dl.min.io/server/minio/hotfixes/linux-${TARGETARCH}/archive/minio.${RELEASE} -o /go/bin/minio && \
+    curl -s -q https://dl.min.io/server/minio/hotfixes/linux-${TARGETARCH}/archive/minio.${RELEASE}.minisig -o /go/bin/minio.minisig && \
+    chmod +x /go/bin/minio
+
+# Download mc binary and signature file
+RUN curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc -o /go/bin/mc && \
+    curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc.minisig -o /go/bin/mc.minisig && \
+    chmod +x /go/bin/mc
+
+# Verify binary signature using public key "RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGavRUN"
+RUN minisign -Vqm /go/bin/minio -x /go/bin/minio.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav && \
+    minisign -Vqm /go/bin/mc -x /go/bin/mc.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav
+
+FROM registry.access.redhat.com/ubi9/ubi-micro:latest
 
 ARG RELEASE
 
@@ -17,33 +44,18 @@ ENV MINIO_ACCESS_KEY_FILE=access_key \
     MINIO_KMS_SECRET_KEY_FILE=kms_master_key \
     MINIO_UPDATE_MINISIGN_PUBKEY="RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav" \
     MINIO_CONFIG_ENV_FILE=config.env \
-    PATH=/opt/bin:$PATH
+    MC_CONFIG_DIR=/tmp/.mc
+
+COPY --from=build /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/
+COPY --from=build /go/bin/minio /usr/bin/minio
+COPY --from=build /go/bin/mc /usr/bin/mc
 
-COPY dockerscripts/verify-minio.sh /usr/bin/verify-minio.sh
-COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 COPY CREDITS /licenses/CREDITS
 COPY LICENSE /licenses/LICENSE
-
-RUN \
-     microdnf clean all && \
-     microdnf update --nodocs && \
-     rpm -Uvh https://dl.fedoraproject.org/pub/epel/epel-release-latest-8.noarch.rpm && \
-     microdnf install curl ca-certificates shadow-utils util-linux gzip lsof tar net-tools iproute iputils jq minisign --nodocs && \
-     mkdir -p /opt/bin && chmod -R 777 /opt/bin && \
-     curl -s -q https://dl.min.io/server/minio/hotfixes/linux-amd64/archive/minio.${RELEASE} -o /opt/bin/minio && \
-     curl -s -q https://dl.min.io/server/minio/hotfixes/linux-amd64/archive/minio.${RELEASE}.sha256sum -o /opt/bin/minio.sha256sum && \
-     curl -s -q https://dl.min.io/server/minio/hotfixes/linux-amd64/archive/minio.${RELEASE}.minisig -o /opt/bin/minio.minisig && \
-     microdnf clean all && \
-     chmod +x /opt/bin/minio && \
-     chmod +x /usr/bin/docker-entrypoint.sh && \
-     chmod +x /usr/bin/verify-minio.sh && \
-     /usr/bin/verify-minio.sh && \
-     microdnf clean all
+COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 
 EXPOSE 9000
-
-ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
-
 VOLUME ["/data"]
 
+ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
 CMD ["minio"]

Dockerfile.release

@@ -1,6 +1,31 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.8
+FROM golang:1.21-alpine as build
 
 ARG TARGETARCH
+ARG RELEASE
+
+ENV GOPATH /go
+ENV CGO_ENABLED 0
+
+# Install curl and minisign
+RUN apk add -U --no-cache ca-certificates && \
+    apk add -U --no-cache curl && \
+    go install aead.dev/minisign/cmd/minisign@v0.2.1
+
+# Download minio binary and signature file
+RUN curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE} -o /go/bin/minio && \
+    curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.minisig -o /go/bin/minio.minisig && \
+    chmod +x /go/bin/minio
+
+# Download mc binary and signature file
+RUN curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc -o /go/bin/mc && \
+    curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc.minisig -o /go/bin/mc.minisig && \
+    chmod +x /go/bin/mc
+
+# Verify binary signature using public key "RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGavRUN"
+RUN minisign -Vqm /go/bin/minio -x /go/bin/minio.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav && \
+    minisign -Vqm /go/bin/mc -x /go/bin/mc.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav
+
+FROM registry.access.redhat.com/ubi9/ubi-micro:latest
 
 ARG RELEASE
 
@@ -19,35 +44,18 @@ ENV MINIO_ACCESS_KEY_FILE=access_key \
     MINIO_KMS_SECRET_KEY_FILE=kms_master_key \
     MINIO_UPDATE_MINISIGN_PUBKEY="RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav" \
     MINIO_CONFIG_ENV_FILE=config.env \
-    PATH=/opt/bin:$PATH
+    MC_CONFIG_DIR=/tmp/.mc
+
+COPY --from=build /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/
+COPY --from=build /go/bin/minio /usr/bin/minio
+COPY --from=build /go/bin/mc /usr/bin/mc
 
-COPY dockerscripts/verify-minio.sh /usr/bin/verify-minio.sh
-COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 COPY CREDITS /licenses/CREDITS
 COPY LICENSE /licenses/LICENSE
-
-RUN \
-     microdnf clean all && \
-     microdnf update --nodocs && \
-     rpm -Uvh https://dl.fedoraproject.org/pub/epel/epel-release-latest-8.noarch.rpm && \
-     microdnf install curl ca-certificates shadow-utils util-linux gzip lsof tar net-tools iproute iputils jq minisign --nodocs && \
-     mkdir -p /opt/bin && chmod -R 777 /opt/bin && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE} -o /opt/bin/minio && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.sha256sum -o /opt/bin/minio.sha256sum && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.minisig -o /opt/bin/minio.minisig && \
-     curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc -o /opt/bin/mc && \
-     microdnf clean all && \
-     chmod +x /opt/bin/minio && \
-     chmod +x /opt/bin/mc && \
-     chmod +x /usr/bin/docker-entrypoint.sh && \
-     chmod +x /usr/bin/verify-minio.sh && \
-     /usr/bin/verify-minio.sh && \
-     microdnf clean all
+COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 
 EXPOSE 9000
-
-ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
-
 VOLUME ["/data"]
 
+ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
 CMD ["minio"]

Dockerfile.release.fips

@@ -1,6 +1,25 @@
-FROM registry.access.redhat.com/ubi8/ubi-minimal:8.8
+FROM golang:1.21-alpine as build
 
 ARG TARGETARCH
+ARG RELEASE
+
+ENV GOPATH /go
+ENV CGO_ENABLED 0
+
+# Install curl and minisign
+RUN apk add -U --no-cache ca-certificates && \
+    apk add -U --no-cache curl && \
+    go install aead.dev/minisign/cmd/minisign@v0.2.1
+
+# Download minio binary and signature file
+RUN curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.fips -o /go/bin/minio && \
+    curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.fips.minisig -o /go/bin/minio.minisig && \
+    chmod +x /go/bin/minio
+
+# Verify binary signature using public key "RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGavRUN"
+RUN minisign -Vqm /go/bin/minio -x /go/bin/minio.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav
+
+FROM registry.access.redhat.com/ubi9/ubi-micro:latest
 
 ARG RELEASE
 
@@ -18,34 +37,17 @@ ENV MINIO_ACCESS_KEY_FILE=access_key \
     MINIO_ROOT_PASSWORD_FILE=secret_key \
     MINIO_KMS_SECRET_KEY_FILE=kms_master_key \
     MINIO_UPDATE_MINISIGN_PUBKEY="RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav" \
-    MINIO_CONFIG_ENV_FILE=config.env \
-    PATH=/opt/bin:$PATH
+    MINIO_CONFIG_ENV_FILE=config.env
+
+COPY --from=build /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/
+COPY --from=build /go/bin/minio /usr/bin/minio
 
-COPY dockerscripts/verify-minio.sh /usr/bin/verify-minio.sh
-COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 COPY CREDITS /licenses/CREDITS
 COPY LICENSE /licenses/LICENSE
-
-RUN \
-     microdnf clean all && \
-     microdnf update --nodocs && \
-     rpm -Uvh https://dl.fedoraproject.org/pub/epel/epel-release-latest-8.noarch.rpm && \
-     microdnf install curl ca-certificates shadow-utils util-linux gzip lsof tar net-tools iproute iputils jq minisign --nodocs && \
-     mkdir -p /opt/bin && chmod -R 777 /opt/bin && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.fips -o /opt/bin/minio && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.fips.sha256sum -o /opt/bin/minio.sha256sum && \
-     curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.fips.minisig -o /opt/bin/minio.minisig && \
-     microdnf clean all && \
-     chmod +x /opt/bin/minio && \
-     chmod +x /usr/bin/docker-entrypoint.sh && \
-     chmod +x /usr/bin/verify-minio.sh && \
-     /usr/bin/verify-minio.sh && \
-     microdnf clean all
+COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
 
 EXPOSE 9000
-
-ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
-
 VOLUME ["/data"]
 
+ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
 CMD ["minio"]

Dockerfile.release.old_cpu

@@ -0,0 +1,61 @@
+FROM golang:1.21-alpine as build
+
+ARG TARGETARCH
+ARG RELEASE
+
+ENV GOPATH /go
+ENV CGO_ENABLED 0
+
+# Install curl and minisign
+RUN apk add -U --no-cache ca-certificates && \
+    apk add -U --no-cache curl && \
+    go install aead.dev/minisign/cmd/minisign@v0.2.1
+
+# Download minio binary and signature file
+RUN curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE} -o /go/bin/minio && \
+    curl -s -q https://dl.min.io/server/minio/release/linux-${TARGETARCH}/archive/minio.${RELEASE}.minisig -o /go/bin/minio.minisig && \
+    chmod +x /go/bin/minio
+
+# Download mc binary and signature file
+RUN curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc -o /go/bin/mc && \
+    curl -s -q https://dl.min.io/client/mc/release/linux-${TARGETARCH}/mc.minisig -o /go/bin/mc.minisig && \
+    chmod +x /go/bin/mc
+
+# Verify binary signature using public key "RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGavRUN"
+RUN minisign -Vqm /go/bin/minio -x /go/bin/minio.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav && \
+    minisign -Vqm /go/bin/mc -x /go/bin/mc.minisig -P RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav
+
+FROM registry.access.redhat.com/ubi8/ubi-micro:latest
+
+ARG RELEASE
+
+LABEL name="MinIO" \
+      vendor="MinIO Inc <dev@min.io>" \
+      maintainer="MinIO Inc <dev@min.io>" \
+      version="${RELEASE}" \
+      release="${RELEASE}" \
+      summary="MinIO is a High Performance Object Storage, API compatible with Amazon S3 cloud storage service." \
+      description="MinIO object storage is fundamentally different. Designed for performance and the S3 API, it is 100% open-source. MinIO is ideal for large, private cloud environments with stringent security requirements and delivers mission-critical availability across a diverse range of workloads."
+
+ENV MINIO_ACCESS_KEY_FILE=access_key \
+    MINIO_SECRET_KEY_FILE=secret_key \
+    MINIO_ROOT_USER_FILE=access_key \
+    MINIO_ROOT_PASSWORD_FILE=secret_key \
+    MINIO_KMS_SECRET_KEY_FILE=kms_master_key \
+    MINIO_UPDATE_MINISIGN_PUBKEY="RWTx5Zr1tiHQLwG9keckT0c45M3AGeHD6IvimQHpyRywVWGbP1aVSGav" \
+    MINIO_CONFIG_ENV_FILE=config.env \
+    MC_CONFIG_DIR=/tmp/.mc
+
+COPY --from=build /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/
+COPY --from=build /go/bin/minio /usr/bin/minio
+COPY --from=build /go/bin/mc /usr/bin/mc
+
+COPY CREDITS /licenses/CREDITS
+COPY LICENSE /licenses/LICENSE
+COPY dockerscripts/docker-entrypoint.sh /usr/bin/docker-entrypoint.sh
+
+EXPOSE 9000
+VOLUME ["/data"]
+
+ENTRYPOINT ["/usr/bin/docker-entrypoint.sh"]
+CMD ["minio"]

Makefile

@@ -6,9 +6,8 @@ GOARCH := $(shell go env GOARCH)
 GOOS := $(shell go env GOOS)
 
 VERSION ?= $(shell git describe --tags)
-TAG ?= "minio/minio:$(VERSION)"
+TAG ?= "quay.io/minio/minio:$(VERSION)"
 
-GOLANGCI_VERSION = v1.51.2
 GOLANGCI_DIR = .bin/golangci/$(GOLANGCI_VERSION)
 GOLANGCI = $(GOLANGCI_DIR)/golangci-lint
 
@@ -23,8 +22,8 @@ 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 $(GOLANGCI_DIR) $(GOLANGCI_VERSION)
-	@echo "Installing msgp" && go install -v github.com/tinylib/msgp@v1.1.7
+	@echo "Installing golangci-lint" && curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/master/install.sh | sh -s -- -b $(GOLANGCI_DIR)
+	@echo "Installing msgp" && go install -v github.com/tinylib/msgp@6ac204f0b4d48d17ab4fa442134c7fba13127a4e
 	@echo "Installing stringer" && go install -v golang.org/x/tools/cmd/stringer@latest
 
 crosscompile: ## cross compile minio
@@ -47,7 +46,7 @@ lint-fix: getdeps ## runs golangci-lint suite of linters with automatic fixes
 check: test
 test: verifiers build build-debugging ## builds minio, runs linters, tests
 	@echo "Running unit tests"
-	@MINIO_API_REQUESTS_MAX=10000 CGO_ENABLED=0 go test -tags kqueue ./...
+	@MINIO_API_REQUESTS_MAX=10000 CGO_ENABLED=0 go test -v -tags kqueue ./...
 
 test-root-disable: install-race
 	@echo "Running minio root lockdown tests"
@@ -60,6 +59,13 @@ test-decom: install-race
 	@env bash $(PWD)/docs/distributed/decom-encrypted-sse-s3.sh
 	@env bash $(PWD)/docs/distributed/decom-compressed-sse-s3.sh
 
+test-versioning: install-race
+	@echo "Running minio versioning tests"
+	@env bash $(PWD)/docs/bucket/versioning/versioning-tests.sh
+
+test-configfile: install-race
+	@env bash $(PWD)/docs/distributed/distributed-from-config-file.sh
+
 test-upgrade: install-race
 	@echo "Running minio upgrade tests"
 	@(env bash $(PWD)/buildscripts/minio-upgrade.sh)
@@ -110,7 +116,6 @@ verify-healing: ## verify healing and replacing disks with minio binary
 	@echo "Verify healing build with race"
 	@GORACE=history_size=7 CGO_ENABLED=1 go build -race -tags kqueue -trimpath --ldflags "$(LDFLAGS)" -o $(PWD)/minio 1>/dev/null
 	@(env bash $(PWD)/buildscripts/verify-healing.sh)
-	@(env bash $(PWD)/buildscripts/unaligned-healing.sh)
 	@(env bash $(PWD)/buildscripts/heal-inconsistent-versions.sh)
 
 verify-healing-with-root-disks: ## verify healing root disks
@@ -141,14 +146,23 @@ hotfix-vars:
 	$(eval VERSION := $(shell git describe --tags --abbrev=0).hotfix.$(shell git rev-parse --short HEAD))
 	$(eval TAG := "minio/minio:$(VERSION)")
 
-hotfix: hotfix-vars install ## builds minio binary with hotfix tags
-	@mv -f ./minio ./minio.$(VERSION)
-	@minisign -qQSm ./minio.$(VERSION) -s "${CRED_DIR}/minisign.key" < "${CRED_DIR}/minisign-passphrase"
-	@sha256sum < ./minio.$(VERSION) | sed 's, -,minio.$(VERSION),g' > minio.$(VERSION).sha256sum
+hotfix: hotfix-vars clean install ## builds minio binary with hotfix tags
+	@wget -q -c https://github.com/minio/pkger/releases/download/v2.2.0/pkger_2.2.0_linux_amd64.deb
+	@wget -q -c https://raw.githubusercontent.com/minio/minio-service/v1.0.1/linux-systemd/distributed/minio.service
+	@sudo apt install ./pkger_2.2.0_linux_amd64.deb --yes
+	@mkdir -p minio-release/$(GOOS)-$(GOARCH)/archive
+	@cp -af ./minio minio-release/$(GOOS)-$(GOARCH)/minio
+	@cp -af ./minio minio-release/$(GOOS)-$(GOARCH)/minio.$(VERSION)
+	@minisign -qQSm minio-release/$(GOOS)-$(GOARCH)/minio.$(VERSION) -s "${CRED_DIR}/minisign.key" < "${CRED_DIR}/minisign-passphrase"
+	@sha256sum < minio-release/$(GOOS)-$(GOARCH)/minio.$(VERSION) | sed 's, -,minio.$(VERSION),g' > minio-release/$(GOOS)-$(GOARCH)/minio.$(VERSION).sha256sum
+	@cp -af minio-release/$(GOOS)-$(GOARCH)/minio.$(VERSION)* minio-release/$(GOOS)-$(GOARCH)/archive/
+	@pkger -r $(VERSION) --ignore
 
 hotfix-push: hotfix
-	@scp -q -r minio.$(VERSION)* minio@dl-0.minio.io:~/releases/server/minio/hotfixes/linux-amd64/archive/
-	@scp -q -r minio.$(VERSION)* minio@dl-1.minio.io:~/releases/server/minio/hotfixes/linux-amd64/archive/
+	@scp -q -r minio-release/$(GOOS)-$(GOARCH)/* minio@dl-0.minio.io:~/releases/server/minio/hotfixes/linux-amd64/
+	@scp -q -r minio-release/$(GOOS)-$(GOARCH)/* minio@dl-0.minio.io:~/releases/server/minio/hotfixes/linux-amd64/archive
+	@scp -q -r minio-release/$(GOOS)-$(GOARCH)/* minio@dl-1.minio.io:~/releases/server/minio/hotfixes/linux-amd64/
+	@scp -q -r minio-release/$(GOOS)-$(GOARCH)/* minio@dl-1.minio.io:~/releases/server/minio/hotfixes/linux-amd64/archive
 	@echo "Published new hotfix binaries at https://dl.min.io/server/minio/hotfixes/linux-amd64/archive/minio.$(VERSION)"
 
 docker-hotfix-push: docker-hotfix
@@ -183,3 +197,6 @@ clean: ## cleanup all generated assets
 	@rm -rvf build
 	@rm -rvf release
 	@rm -rvf .verify*
+	@rm -rvf minio-release
+	@rm -rvf minio.RELEASE*.hotfix.*
+	@rm -rvf pkger_*.deb

buildscripts/cross-compile.sh

@@ -9,7 +9,7 @@ function _init() {
 	export CGO_ENABLED=0
 
 	## List of architectures and OS to test coss compilation.
-	SUPPORTED_OSARCH="linux/ppc64le linux/mips64 linux/amd64 linux/arm64 linux/s390x darwin/arm64 darwin/amd64 freebsd/amd64 windows/amd64 linux/arm linux/386 netbsd/amd64 linux/mips"
+	SUPPORTED_OSARCH="linux/ppc64le linux/mips64 linux/amd64 linux/arm64 linux/s390x darwin/arm64 darwin/amd64 freebsd/amd64 windows/amd64 linux/arm linux/386 netbsd/amd64 linux/mips openbsd/amd64"
 }
 
 function _build() {

buildscripts/rewrite-old-new.sh

@@ -87,8 +87,8 @@ function verify_rewrite() {
 		exit 1
 	fi
 
-	go build ./docs/debugging/s3-check-md5/
-	if ! ./s3-check-md5 \
+	go install github.com/minio/minio/docs/debugging/s3-check-md5@master
+	if ! s3-check-md5 \
 		-debug \
 		-versions \
 		-access-key minio \
@@ -113,7 +113,7 @@ function verify_rewrite() {
 	go run ./buildscripts/heal-manual.go "127.0.0.1:${start_port}" "minio" "minio123"
 	sleep 1
 
-	if ! ./s3-check-md5 \
+	if ! s3-check-md5 \
 		-debug \
 		-versions \
 		-access-key minio \

buildscripts/unaligned-healing.sh

@@ -1,177 +0,0 @@
-#!/bin/bash -e
-#
-
-set -E
-set -o pipefail
-set -x
-
-if [ ! -x "$PWD/minio" ]; then
-	echo "minio executable binary not found in current directory"
-	exit 1
-fi
-
-WORK_DIR="$PWD/.verify-$RANDOM"
-MINIO_CONFIG_DIR="$WORK_DIR/.minio"
-MINIO_OLD=("$PWD/minio.RELEASE.2021-11-24T23-19-33Z" --config-dir "$MINIO_CONFIG_DIR" server)
-MINIO=("$PWD/minio" --config-dir "$MINIO_CONFIG_DIR" server)
-
-function download_old_release() {
-	if [ ! -f minio.RELEASE.2021-11-24T23-19-33Z ]; then
-		curl --silent -O https://dl.minio.io/server/minio/release/linux-amd64/archive/minio.RELEASE.2021-11-24T23-19-33Z
-		chmod a+x minio.RELEASE.2021-11-24T23-19-33Z
-	fi
-}
-
-function start_minio_16drive() {
-	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_SHARD_DISKTIME_DELTA="5s" # do not change this as its needed for tests
-	export MINIO_CI_CD=1
-
-	MC_BUILD_DIR="mc-$RANDOM"
-	if ! git clone --quiet https://github.com/minio/mc "$MC_BUILD_DIR"; then
-		echo "failed to download https://github.com/minio/mc"
-		purge "${MC_BUILD_DIR}"
-		exit 1
-	fi
-
-	(cd "${MC_BUILD_DIR}" && go build -o "$WORK_DIR/mc")
-
-	# remove mc source.
-	purge "${MC_BUILD_DIR}"
-
-	"${MINIO_OLD[@]}" --address ":$start_port" "${WORK_DIR}/xl{1...16}" >"${WORK_DIR}/server1.log" 2>&1 &
-	pid=$!
-	disown $pid
-	sleep 30
-
-	if ! ps -p ${pid} 1>&2 >/dev/null; then
-		echo "server1 log:"
-		cat "${WORK_DIR}/server1.log"
-		echo "FAILED"
-		purge "$WORK_DIR"
-		exit 1
-	fi
-
-	shred --iterations=1 --size=5241856 - 1>"${WORK_DIR}/unaligned" 2>/dev/null
-	"${WORK_DIR}/mc" mb minio/healing-shard-bucket --quiet
-	"${WORK_DIR}/mc" cp \
-		"${WORK_DIR}/unaligned" \
-		minio/healing-shard-bucket/unaligned \
-		--disable-multipart --quiet
-
-	## "unaligned" object name gets consistently distributed
-	## to disks in following distribution order
-	##
-	## NOTE: if you change the name make sure to change the
-	## distribution order present here
-	##
-	## [15, 16, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]
-
-	## make sure to remove the "last" data shard
-	rm -rf "${WORK_DIR}/xl14/healing-shard-bucket/unaligned"
-	sleep 10
-	## Heal the shard
-	"${WORK_DIR}/mc" admin heal --quiet --recursive minio/healing-shard-bucket
-	## then remove any other data shard let's pick first disk
-	## - 1st data shard.
-	rm -rf "${WORK_DIR}/xl3/healing-shard-bucket/unaligned"
-	sleep 10
-
-	go build ./docs/debugging/s3-check-md5/
-	if ! ./s3-check-md5 \
-		-debug \
-		-access-key minio \
-		-secret-key minio123 \
-		-endpoint http://127.0.0.1:${start_port}/ 2>&1 | grep CORRUPTED; then
-		echo "server1 log:"
-		cat "${WORK_DIR}/server1.log"
-		echo "FAILED"
-		purge "$WORK_DIR"
-		exit 1
-	fi
-
-	pkill minio
-	sleep 3
-
-	"${MINIO[@]}" --address ":$start_port" "${WORK_DIR}/xl{1...16}" >"${WORK_DIR}/server1.log" 2>&1 &
-	pid=$!
-	disown $pid
-	sleep 30
-
-	if ! ps -p ${pid} 1>&2 >/dev/null; then
-		echo "server1 log:"
-		cat "${WORK_DIR}/server1.log"
-		echo "FAILED"
-		purge "$WORK_DIR"
-		exit 1
-	fi
-
-	if ! ./s3-check-md5 \
-		-debug \
-		-access-key minio \
-		-secret-key minio123 \
-		-endpoint http://127.0.0.1:${start_port}/ 2>&1 | grep INTACT; then
-		echo "server1 log:"
-		cat "${WORK_DIR}/server1.log"
-		echo "FAILED"
-		mkdir -p inspects
-		(
-			cd inspects
-			"${WORK_DIR}/mc" support inspect minio/healing-shard-bucket/unaligned/**
-		)
-
-		"${WORK_DIR}/mc" mb play/inspects
-		"${WORK_DIR}/mc" mirror inspects play/inspects
-
-		purge "$WORK_DIR"
-		exit 1
-	fi
-
-	"${WORK_DIR}/mc" admin heal --quiet --recursive minio/healing-shard-bucket
-
-	if ! ./s3-check-md5 \
-		-debug \
-		-access-key minio \
-		-secret-key minio123 \
-		-endpoint http://127.0.0.1:${start_port}/ 2>&1 | grep INTACT; then
-		echo "server1 log:"
-		cat "${WORK_DIR}/server1.log"
-		echo "FAILED"
-		mkdir -p inspects
-		(
-			cd inspects
-			"${WORK_DIR}/mc" support inspect minio/healing-shard-bucket/unaligned/**
-		)
-
-		"${WORK_DIR}/mc" mb play/inspects
-		"${WORK_DIR}/mc" mirror inspects play/inspects
-
-		purge "$WORK_DIR"
-		exit 1
-	fi
-
-	pkill minio
-	sleep 3
-}
-
-function main() {
-	download_old_release
-
-	start_port=$(shuf -i 10000-65000 -n 1)
-
-	start_minio_16drive ${start_port}
-}
-
-function purge() {
-	rm -rf "$1"
-}
-
-(main "$@")
-rv=$?
-purge "$WORK_DIR"
-exit "$rv"

buildscripts/upgrade-tests/compose.yml

@@ -9,11 +9,6 @@ x-minio-common: &minio-common
   expose:
     - "9000"
     - "9001"
-  healthcheck:
-    test: ["CMD", "curl", "-f", "http://localhost:9000/minio/health/live"]
-    interval: 30s
-    timeout: 20s
-    retries: 3
 
 # starts 4 docker containers running minio server instances.
 # using nginx reverse proxy, load balancing, you can access

cmd/admin-bucket-handlers.go

@@ -94,7 +94,7 @@ func (a adminAPIHandlers) PutBucketQuotaConfigHandler(w http.ResponseWriter, r *
 		Quota:     data,
 		UpdatedAt: updatedAt,
 	}
-	if quotaConfig.Size == 0 || quotaConfig.Quota == 0 {
+	if quotaConfig.Size == 0 && quotaConfig.Quota == 0 {
 		bucketMeta.Quota = nil
 	}
 
@@ -392,7 +392,8 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 	// of bucket metadata
 	zipWriter := zip.NewWriter(w)
 	defer zipWriter.Close()
-	rawDataFn := func(r io.Reader, filename string, sz int) error {
+
+	rawDataFn := func(r io.Reader, filename string, sz int) {
 		header, zerr := zip.FileInfoHeader(dummyFileInfo{
 			name:    filename,
 			size:    int64(sz),
@@ -401,20 +402,13 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 			isDir:   false,
 			sys:     nil,
 		})
-		if zerr != nil {
-			logger.LogIf(ctx, zerr)
-			return nil
+		if zerr == nil {
+			header.Method = zip.Deflate
+			zwriter, zerr := zipWriter.CreateHeader(header)
+			if zerr == nil {
+				io.Copy(zwriter, r)
+			}
 		}
-		header.Method = zip.Deflate
-		zwriter, zerr := zipWriter.CreateHeader(header)
-		if zerr != nil {
-			logger.LogIf(ctx, zerr)
-			return nil
-		}
-		if _, err := io.Copy(zwriter, r); err != nil {
-			logger.LogIf(ctx, err)
-		}
-		return nil
 	}
 
 	cfgFiles := []string{
@@ -446,10 +440,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketLifecycleConfig:
 				config, _, err := globalBucketMetadataSys.GetLifecycleConfig(bucket)
 				if err != nil {
@@ -465,10 +456,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketQuotaConfigFile:
 				config, _, err := globalBucketMetadataSys.GetQuotaConfig(ctx, bucket)
 				if err != nil {
@@ -483,10 +471,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketSSEConfig:
 				config, _, err := globalBucketMetadataSys.GetSSEConfig(bucket)
 				if err != nil {
@@ -501,10 +486,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketTaggingConfig:
 				config, _, err := globalBucketMetadataSys.GetTaggingConfig(bucket)
 				if err != nil {
@@ -519,10 +501,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case objectLockConfig:
 				config, _, err := globalBucketMetadataSys.GetObjectLockConfig(bucket)
 				if err != nil {
@@ -538,10 +517,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketVersioningConfig:
 				config, _, err := globalBucketMetadataSys.GetVersioningConfig(bucket)
 				if err != nil {
@@ -557,10 +533,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketReplicationConfig:
 				config, _, err := globalBucketMetadataSys.GetReplicationConfig(ctx, bucket)
 				if err != nil {
@@ -575,11 +548,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			case bucketTargetsFile:
 				config, err := globalBucketMetadataSys.GetBucketTargetsConfig(bucket)
 				if err != nil {
@@ -595,10 +564,7 @@ func (a adminAPIHandlers) ExportBucketMetadataHandler(w http.ResponseWriter, r *
 					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
 					return
 				}
-				if err = rawDataFn(bytes.NewReader(configData), cfgPath, len(configData)); err != nil {
-					writeErrorResponse(ctx, w, exportError(ctx, err, cfgFile, bucket), r.URL)
-					return
-				}
+				rawDataFn(bytes.NewReader(configData), cfgPath, len(configData))
 			}
 		}
 	}
@@ -714,24 +680,17 @@ func (a adminAPIHandlers) ImportBucketMetadataHandler(w http.ResponseWriter, r *
 			if _, ok := bucketMap[bucket]; !ok {
 				opts := MakeBucketOptions{
 					LockEnabled: config.Enabled(),
+					ForceCreate: true, // ignore if it already exists
 				}
 				err = objectAPI.MakeBucket(ctx, bucket, opts)
 				if err != nil {
-					if _, ok := err.(BucketExists); !ok {
-						rpt.SetStatus(bucket, fileName, err)
-						continue
-					}
+					rpt.SetStatus(bucket, fileName, err)
+					continue
 				}
 				v := newBucketMetadata(bucket)
 				bucketMap[bucket] = &v
 			}
 
-			// Deny object locking configuration settings on existing buckets without object lock enabled.
-			if _, _, err = globalBucketMetadataSys.GetObjectLockConfig(bucket); err != nil {
-				rpt.SetStatus(bucket, fileName, err)
-				continue
-			}
-
 			bucketMap[bucket].ObjectLockConfigXML = configData
 			bucketMap[bucket].ObjectLockConfigUpdatedAt = updatedAt
 			rpt.SetStatus(bucket, fileName, nil)
@@ -758,11 +717,11 @@ func (a adminAPIHandlers) ImportBucketMetadataHandler(w http.ResponseWriter, r *
 				continue
 			}
 			if _, ok := bucketMap[bucket]; !ok {
-				if err = objectAPI.MakeBucket(ctx, bucket, MakeBucketOptions{}); err != nil {
-					if _, ok := err.(BucketExists); !ok {
-						rpt.SetStatus(bucket, fileName, err)
-						continue
-					}
+				if err = objectAPI.MakeBucket(ctx, bucket, MakeBucketOptions{
+					ForceCreate: true, // ignore if it already exists
+				}); err != nil {
+					rpt.SetStatus(bucket, fileName, err)
+					continue
 				}
 				v := newBucketMetadata(bucket)
 				bucketMap[bucket] = &v
@@ -810,12 +769,12 @@ func (a adminAPIHandlers) ImportBucketMetadataHandler(w http.ResponseWriter, r *
 
 		// create bucket if it does not exist yet.
 		if _, ok := bucketMap[bucket]; !ok {
-			err = objectAPI.MakeBucket(ctx, bucket, MakeBucketOptions{})
+			err = objectAPI.MakeBucket(ctx, bucket, MakeBucketOptions{
+				ForceCreate: true, // ignore if it already exists
+			})
 			if err != nil {
-				if _, ok := err.(BucketExists); !ok {
-					rpt.SetStatus(bucket, "", err)
-					continue
-				}
+				rpt.SetStatus(bucket, "", err)
+				continue
 			}
 			v := newBucketMetadata(bucket)
 			bucketMap[bucket] = &v

cmd/admin-handlers-config-kv.go

@@ -28,7 +28,6 @@ import (
 
 	"github.com/minio/madmin-go/v3"
 	"github.com/minio/minio/internal/config"
-	"github.com/minio/minio/internal/config/cache"
 	"github.com/minio/minio/internal/config/etcd"
 	xldap "github.com/minio/minio/internal/config/identity/ldap"
 	"github.com/minio/minio/internal/config/identity/openid"
@@ -59,7 +58,7 @@ func (a adminAPIHandlers) DelConfigKVHandler(w http.ResponseWriter, r *http.Requ
 	password := cred.SecretKey
 	kvBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
-		logger.LogIf(ctx, err, logger.Application)
+		logger.LogIf(ctx, err, logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
 		return
 	}
@@ -163,7 +162,7 @@ func (a adminAPIHandlers) SetConfigKVHandler(w http.ResponseWriter, r *http.Requ
 	password := cred.SecretKey
 	kvBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
-		logger.LogIf(ctx, err, logger.Application)
+		logger.LogIf(ctx, err, logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
 		return
 	}
@@ -444,7 +443,7 @@ func (a adminAPIHandlers) SetConfigHandler(w http.ResponseWriter, r *http.Reques
 	password := cred.SecretKey
 	kvBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
-		logger.LogIf(ctx, err, logger.Application)
+		logger.LogIf(ctx, err, logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
 		return
 	}
@@ -500,8 +499,6 @@ func (a adminAPIHandlers) GetConfigHandler(w http.ResponseWriter, r *http.Reques
 			switch hkv.Key {
 			case config.EtcdSubSys:
 				off = !etcd.Enabled(item.Config)
-			case config.CacheSubSys:
-				off = !cache.Enabled(item.Config)
 			case config.StorageClassSubSys:
 				off = !storageclass.Enabled(item.Config)
 			case config.PolicyPluginSubSys:

cmd/admin-handlers-idp-config.go

@@ -60,7 +60,7 @@ func addOrUpdateIDPHandler(ctx context.Context, w http.ResponseWriter, r *http.R
 	password := cred.SecretKey
 	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
-		logger.LogIf(ctx, err, logger.Application)
+		logger.LogIf(ctx, err, logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
 		return
 	}
@@ -121,7 +121,7 @@ func addOrUpdateIDPHandler(ctx context.Context, w http.ResponseWriter, r *http.R
 
 	// IDP config is not dynamic. Sanity check.
 	if dynamic {
-		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInternalError), err.Error(), r.URL)
+		writeCustomErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrInternalError), "", r.URL)
 		return
 	}
 

cmd/admin-handlers-idp-ldap.go

@@ -19,10 +19,14 @@ package cmd
 
 import (
 	"encoding/json"
+	"errors"
+	"fmt"
 	"io"
 	"net/http"
+	"strings"
 
 	"github.com/minio/madmin-go/v3"
+	"github.com/minio/minio/internal/auth"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/mux"
 	"github.com/minio/pkg/v2/policy"
@@ -128,7 +132,7 @@ func (a adminAPIHandlers) AttachDetachPolicyLDAP(w http.ResponseWriter, r *http.
 	password := cred.SecretKey
 	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
-		logger.LogIf(ctx, err, logger.Application)
+		logger.LogIf(ctx, err, logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminConfigBadJSON), r.URL)
 		return
 	}
@@ -175,3 +179,267 @@ func (a adminAPIHandlers) AttachDetachPolicyLDAP(w http.ResponseWriter, r *http.
 
 	writeSuccessResponseJSON(w, encryptedData)
 }
+
+// AddServiceAccountLDAP adds a new service account for provided LDAP username or DN
+//
+// PUT /minio/admin/v3/idp/ldap/add-service-account
+func (a adminAPIHandlers) AddServiceAccountLDAP(w http.ResponseWriter, r *http.Request) {
+	ctx, cred, opts, createReq, targetUser, APIError := commonAddServiceAccount(r)
+	if APIError.Code != "" {
+		writeErrorResponseJSON(ctx, w, APIError, r.URL)
+		return
+	}
+
+	// fail if ldap is not enabled
+	if !globalIAMSys.LDAPConfig.Enabled() {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errors.New("LDAP not enabled")), r.URL)
+	}
+
+	// Find the user for the request sender (as it may be sent via a service
+	// account or STS account):
+	requestorUser := cred.AccessKey
+	requestorParentUser := cred.AccessKey
+	requestorGroups := cred.Groups
+	requestorIsDerivedCredential := false
+	if cred.IsServiceAccount() || cred.IsTemp() {
+		requestorParentUser = cred.ParentUser
+		requestorIsDerivedCredential = true
+	}
+
+	// Check if we are creating svc account for request sender.
+	isSvcAccForRequestor := false
+	if targetUser == requestorUser || targetUser == requestorParentUser {
+		isSvcAccForRequestor = true
+	}
+
+	var (
+		targetGroups []string
+		err          error
+	)
+
+	// If we are creating svc account for request sender, ensure
+	// that targetUser is a real user (i.e. not derived
+	// credentials).
+	if isSvcAccForRequestor {
+		if requestorIsDerivedCredential {
+			if requestorParentUser == "" {
+				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx,
+					errors.New("service accounts cannot be generated for temporary credentials without parent")), r.URL)
+				return
+			}
+			targetUser = requestorParentUser
+		}
+		targetGroups = requestorGroups
+
+		// Deny if the target user is not LDAP
+		isLDAP, err := globalIAMSys.LDAPConfig.DoesUsernameExist(targetUser)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+		if isLDAP == "" {
+			err := errors.New("Specified user does not exist on LDAP server")
+			APIErr := errorCodes.ToAPIErrWithErr(ErrAdminNoSuchUser, err)
+			writeErrorResponseJSON(ctx, w, APIErr, r.URL)
+			return
+		}
+
+		// In case of LDAP/OIDC we need to set `opts.claims` to ensure
+		// it is associated with the LDAP/OIDC user properly.
+		for k, v := range cred.Claims {
+			if k == expClaim {
+				continue
+			}
+			opts.claims[k] = v
+		}
+	} else {
+		isDN := globalIAMSys.LDAPConfig.IsLDAPUserDN(targetUser)
+
+		opts.claims[ldapUserN] = targetUser // simple username
+		targetUser, targetGroups, err = globalIAMSys.LDAPConfig.LookupUserDN(targetUser)
+		if err != nil {
+			// if not found, check if DN
+			if strings.Contains(err.Error(), "not found") && isDN {
+				// warn user that DNs are not allowed
+				err = fmt.Errorf("Must use short username to add service account. %w", err)
+			}
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+		opts.claims[ldapUser] = targetUser // DN
+	}
+
+	newCred, updatedAt, err := globalIAMSys.NewServiceAccount(ctx, targetUser, targetGroups, opts)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	createResp := madmin.AddServiceAccountResp{
+		Credentials: madmin.Credentials{
+			AccessKey:  newCred.AccessKey,
+			SecretKey:  newCred.SecretKey,
+			Expiration: newCred.Expiration,
+		},
+	}
+
+	data, err := json.Marshal(createResp)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	encryptedData, err := madmin.EncryptData(cred.SecretKey, data)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, encryptedData)
+
+	// Call hook for cluster-replication if the service account is not for a
+	// root user.
+	if newCred.ParentUser != globalActiveCred.AccessKey {
+		logger.LogIf(ctx, 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,
+					Name:          newCred.Name,
+					Description:   newCred.Description,
+					Claims:        opts.claims,
+					SessionPolicy: createReq.Policy,
+					Status:        auth.AccountOn,
+					Expiration:    createReq.Expiration,
+				},
+			},
+			UpdatedAt: updatedAt,
+		}))
+	}
+}
+
+// ListAccessKeysLDAP - GET /minio/admin/v3/idp/ldap/list-access-keys
+func (a adminAPIHandlers) ListAccessKeysLDAP(w http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	// Get current object layer instance.
+	objectAPI := newObjectLayerFn()
+	if objectAPI == nil || globalNotificationSys == nil {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
+		return
+	}
+
+	cred, owner, s3Err := validateAdminSignature(ctx, r, "")
+	if s3Err != ErrNone {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(s3Err), r.URL)
+		return
+	}
+
+	userDN := r.Form.Get("userDN")
+
+	// If listing is requested for a specific user (who is not the request
+	// sender), check that the user has permissions.
+	if userDN != "" && userDN != cred.ParentUser {
+		if !globalIAMSys.IsAllowed(policy.Args{
+			AccountName:     cred.AccessKey,
+			Groups:          cred.Groups,
+			Action:          policy.ListServiceAccountsAdminAction,
+			ConditionValues: getConditionValues(r, "", cred),
+			IsOwner:         owner,
+			Claims:          cred.Claims,
+		}) {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
+			return
+		}
+	} else {
+		if !globalIAMSys.IsAllowed(policy.Args{
+			AccountName:     cred.AccessKey,
+			Groups:          cred.Groups,
+			Action:          policy.ListServiceAccountsAdminAction,
+			ConditionValues: getConditionValues(r, "", cred),
+			IsOwner:         owner,
+			Claims:          cred.Claims,
+			DenyOnly:        true,
+		}) {
+			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
+			return
+		}
+		userDN = cred.AccessKey
+		if cred.ParentUser != "" {
+			userDN = cred.ParentUser
+		}
+	}
+
+	targetAccount, err := globalIAMSys.LDAPConfig.DoesUsernameExist(userDN)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	} else if userDN == "" {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errNoSuchUser), r.URL)
+		return
+	}
+
+	listType := r.Form.Get("listType")
+	if listType != "sts-only" && listType != "svcacc-only" && listType != "" {
+		// default to both
+		listType = ""
+	}
+
+	var serviceAccounts []auth.Credentials
+	var stsKeys []auth.Credentials
+
+	if listType == "" || listType == "sts-only" {
+		stsKeys, err = globalIAMSys.ListSTSAccounts(ctx, targetAccount)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+	}
+	if listType == "" || listType == "svcacc-only" {
+		serviceAccounts, err = globalIAMSys.ListServiceAccounts(ctx, targetAccount)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
+		}
+	}
+
+	var serviceAccountList []madmin.ServiceAccountInfo
+	var stsKeyList []madmin.ServiceAccountInfo
+
+	for _, svc := range serviceAccounts {
+		expiryTime := svc.Expiration
+		serviceAccountList = append(serviceAccountList, madmin.ServiceAccountInfo{
+			AccessKey:  svc.AccessKey,
+			Expiration: &expiryTime,
+		})
+	}
+	for _, sts := range stsKeys {
+		expiryTime := sts.Expiration
+		stsKeyList = append(stsKeyList, madmin.ServiceAccountInfo{
+			AccessKey:  sts.AccessKey,
+			Expiration: &expiryTime,
+		})
+	}
+
+	listResp := madmin.ListAccessKeysLDAPResp{
+		ServiceAccounts: serviceAccountList,
+		STSKeys:         stsKeyList,
+	}
+
+	data, err := json.Marshal(listResp)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	encryptedData, err := madmin.EncryptData(cred.SecretKey, data)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, encryptedData)
+}

cmd/admin-handlers-pools.go

@@ -22,6 +22,7 @@ import (
 	"errors"
 	"fmt"
 	"net/http"
+	"strconv"
 	"strings"
 
 	"github.com/minio/minio/internal/logger"
@@ -66,16 +67,28 @@ func (a adminAPIHandlers) StartDecommission(w http.ResponseWriter, r *http.Reque
 
 	vars := mux.Vars(r)
 	v := vars["pool"]
+	byID := vars["by-id"] == "true"
 
 	pools := strings.Split(v, ",")
 	poolIndices := make([]int, 0, len(pools))
 
 	for _, pool := range pools {
-		idx := globalEndpoints.GetPoolIdx(pool)
-		if idx == -1 {
-			// We didn't find any matching pools, invalid input
-			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errInvalidArgument), r.URL)
-			return
+		var idx int
+		if byID {
+			var err error
+			idx, err = strconv.Atoi(pool)
+			if err != nil {
+				// We didn't find any matching pools, invalid input
+				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errInvalidArgument), r.URL)
+				return
+			}
+		} else {
+			idx = globalEndpoints.GetPoolIdx(pool)
+			if idx == -1 {
+				// We didn't find any matching pools, invalid input
+				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errInvalidArgument), r.URL)
+				return
+			}
 		}
 		var pool *erasureSets
 		for pidx := range z.serverPools {
@@ -132,8 +145,17 @@ func (a adminAPIHandlers) CancelDecommission(w http.ResponseWriter, r *http.Requ
 
 	vars := mux.Vars(r)
 	v := vars["pool"]
+	byID := vars["by-id"] == "true"
+	idx := -1
+
+	if byID {
+		if i, err := strconv.Atoi(v); err == nil && i >= 0 && i < len(globalEndpoints) {
+			idx = i
+		}
+	} else {
+		idx = globalEndpoints.GetPoolIdx(v)
+	}
 
-	idx := globalEndpoints.GetPoolIdx(v)
 	if idx == -1 {
 		// We didn't find any matching pools, invalid input
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, errInvalidArgument), r.URL)
@@ -178,8 +200,17 @@ func (a adminAPIHandlers) StatusPool(w http.ResponseWriter, r *http.Request) {
 
 	vars := mux.Vars(r)
 	v := vars["pool"]
+	byID := vars["by-id"] == "true"
+	idx := -1
+
+	if byID {
+		if i, err := strconv.Atoi(v); err == nil && i >= 0 && i < len(globalEndpoints) {
+			idx = i
+		}
+	} else {
+		idx = globalEndpoints.GetPoolIdx(v)
+	}
 
-	idx := globalEndpoints.GetPoolIdx(v)
 	if idx == -1 {
 		apiErr := toAdminAPIErr(ctx, errInvalidArgument)
 		apiErr.Description = fmt.Sprintf("specified pool '%s' not found, please specify a valid pool", v)

cmd/admin-handlers-site-replication.go

@@ -31,6 +31,7 @@ import (
 
 	"github.com/dustin/go-humanize"
 	"github.com/minio/madmin-go/v3"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/mux"
 	"github.com/minio/pkg/v2/policy"
@@ -51,7 +52,8 @@ func (a adminAPIHandlers) SiteReplicationAdd(w http.ResponseWriter, r *http.Requ
 		return
 	}
 
-	status, err := globalSiteReplicationSys.AddPeerClusters(ctx, sites)
+	opts := getSRAddOptions(r)
+	status, err := globalSiteReplicationSys.AddPeerClusters(ctx, sites, opts)
 	if err != nil {
 		logger.LogIf(ctx, err)
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -67,6 +69,11 @@ func (a adminAPIHandlers) SiteReplicationAdd(w http.ResponseWriter, r *http.Requ
 	writeSuccessResponseJSON(w, body)
 }
 
+func getSRAddOptions(r *http.Request) (opts madmin.SRAddOptions) {
+	opts.ReplicateILMExpiry = r.Form.Get("replicateILMExpiry") == "true"
+	return
+}
+
 // SRPeerJoin - PUT /minio/admin/v3/site-replication/join
 //
 // used internally to tell current cluster to enable SR with
@@ -191,7 +198,7 @@ func (a adminAPIHandlers) SRPeerReplicateIAMItem(w http.ResponseWriter, r *http.
 	}
 }
 
-// SRPeerReplicateBucketItem - PUT /minio/admin/v3/site-replication/bucket-meta
+// SRPeerReplicateBucketItem - PUT /minio/admin/v3/site-replication/peer/bucket-meta
 func (a adminAPIHandlers) SRPeerReplicateBucketItem(w http.ResponseWriter, r *http.Request) {
 	ctx := r.Context()
 
@@ -252,6 +259,8 @@ func (a adminAPIHandlers) SRPeerReplicateBucketItem(w http.ResponseWriter, r *ht
 		err = globalSiteReplicationSys.PeerBucketObjectLockConfigHandler(ctx, item.Bucket, item.ObjectLockConfig, item.UpdatedAt)
 	case madmin.SRBucketMetaTypeSSEConfig:
 		err = globalSiteReplicationSys.PeerBucketSSEConfigHandler(ctx, item.Bucket, item.SSEConfig, item.UpdatedAt)
+	case madmin.SRBucketMetaLCConfig:
+		err = globalSiteReplicationSys.PeerBucketLCConfigHandler(ctx, item.Bucket, item.ExpiryLCConfig, item.UpdatedAt)
 	}
 	if err != nil {
 		logger.LogIf(ctx, err)
@@ -333,6 +342,7 @@ func (a adminAPIHandlers) SiteReplicationStatus(w http.ResponseWriter, r *http.R
 		opts.Users = true
 		opts.Policies = true
 		opts.Groups = true
+		opts.ILMExpiryRules = true
 	}
 	info, err := globalSiteReplicationSys.SiteReplicationStatus(ctx, objectAPI, opts)
 	if err != nil {
@@ -382,7 +392,9 @@ func (a adminAPIHandlers) SiteReplicationEdit(w http.ResponseWriter, r *http.Req
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
 	}
-	status, err := globalSiteReplicationSys.EditPeerCluster(ctx, site)
+
+	opts := getSREditOptions(r)
+	status, err := globalSiteReplicationSys.EditPeerCluster(ctx, site, opts)
 	if err != nil {
 		logger.LogIf(ctx, err)
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -397,6 +409,12 @@ func (a adminAPIHandlers) SiteReplicationEdit(w http.ResponseWriter, r *http.Req
 	writeSuccessResponseJSON(w, body)
 }
 
+func getSREditOptions(r *http.Request) (opts madmin.SREditOptions) {
+	opts.DisableILMExpiryReplication = r.Form.Get("disableILMExpiryReplication") == "true"
+	opts.EnableILMExpiryReplication = r.Form.Get("enableILMExpiryReplication") == "true"
+	return
+}
+
 // SRPeerEdit - PUT /minio/admin/v3/site-replication/peer/edit
 //
 // used internally to tell current cluster to update endpoint for peer
@@ -421,12 +439,37 @@ func (a adminAPIHandlers) SRPeerEdit(w http.ResponseWriter, r *http.Request) {
 	}
 }
 
+// SRStateEdit - PUT /minio/admin/v3/site-replication/state/edit
+//
+// used internally to tell current cluster to update site replication state
+func (a adminAPIHandlers) SRStateEdit(w http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	objectAPI, _ := validateAdminReq(ctx, w, r, policy.SiteReplicationOperationAction)
+	if objectAPI == nil {
+		return
+	}
+
+	var state madmin.SRStateEditReq
+	if err := parseJSONBody(ctx, r.Body, &state, ""); err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+	if err := globalSiteReplicationSys.PeerStateEditReq(ctx, state); err != nil {
+		logger.LogIf(ctx, err)
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+}
+
 func getSRStatusOptions(r *http.Request) (opts madmin.SRStatusOptions) {
 	q := r.Form
 	opts.Buckets = q.Get("buckets") == "true"
 	opts.Policies = q.Get("policies") == "true"
 	opts.Groups = q.Get("groups") == "true"
 	opts.Users = q.Get("users") == "true"
+	opts.ILMExpiryRules = q.Get("ilm-expiry-rules") == "true"
+	opts.PeerState = q.Get("peer-state") == "true"
 	opts.Entity = madmin.GetSREntityType(q.Get("entity"))
 	opts.EntityValue = q.Get("entityvalue")
 	opts.ShowDeleted = q.Get("showDeleted") == "true"
@@ -537,7 +580,7 @@ func (a adminAPIHandlers) SiteReplicationDevNull(w http.ResponseWriter, r *http.
 
 	connectTime := time.Now()
 	for {
-		n, err := io.CopyN(io.Discard, r.Body, 128*humanize.KiByte)
+		n, err := io.CopyN(xioutil.Discard, r.Body, 128*humanize.KiByte)
 		atomic.AddUint64(&globalSiteNetPerfRX.RX, uint64(n))
 		if err != nil && err != io.EOF && err != io.ErrUnexpectedEOF {
 			// If there is a disconnection before globalNetPerfMinDuration (we give a margin of error of 1 sec)

cmd/admin-handlers-users.go

@@ -27,6 +27,7 @@ import (
 	"net/http"
 	"os"
 	"sort"
+	"strconv"
 	"time"
 
 	"github.com/klauspost/compress/zip"
@@ -545,14 +546,16 @@ func (a adminAPIHandlers) TemporaryAccountInfo(w http.ResponseWriter, r *http.Re
 		return
 	}
 
-	if !globalIAMSys.IsAllowed(policy.Args{
+	args := policy.Args{
 		AccountName:     cred.AccessKey,
 		Groups:          cred.Groups,
 		Action:          policy.ListTemporaryAccountsAdminAction,
 		ConditionValues: getConditionValues(r, "", cred),
 		IsOwner:         owner,
 		Claims:          cred.Claims,
-	}) {
+	}
+
+	if !globalIAMSys.IsAllowed(args) {
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
 		return
 	}
@@ -568,11 +571,16 @@ func (a adminAPIHandlers) TemporaryAccountInfo(w http.ResponseWriter, r *http.Re
 	if sessionPolicy != nil {
 		stsAccountPolicy = *sessionPolicy
 	} else {
-		policiesNames, err := globalIAMSys.PolicyDBGet(stsAccount.ParentUser, false)
+		policiesNames, err := globalIAMSys.PolicyDBGet(stsAccount.ParentUser, stsAccount.Groups...)
 		if err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
 		}
+		if len(policiesNames) == 0 {
+			policySet, _ := args.GetPolicies(iamPolicyClaimNameOpenID())
+			policiesNames = policySet.ToSlice()
+		}
+
 		stsAccountPolicy = globalIAMSys.GetCombinedPolicy(policiesNames...)
 	}
 
@@ -607,72 +615,17 @@ func (a adminAPIHandlers) TemporaryAccountInfo(w http.ResponseWriter, r *http.Re
 
 // AddServiceAccount - PUT /minio/admin/v3/add-service-account
 func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Request) {
-	ctx := r.Context()
-
-	// Get current object layer instance.
-	objectAPI := newObjectLayerFn()
-	if objectAPI == nil || globalNotificationSys == nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL)
-		return
-	}
-
-	cred, owner, s3Err := validateAdminSignature(ctx, r, "")
-	if s3Err != ErrNone {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(s3Err), r.URL)
-		return
-	}
-
-	password := cred.SecretKey
-	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err), r.URL)
-		return
-	}
-
-	var createReq madmin.AddServiceAccountReq
-	if err = json.Unmarshal(reqBytes, &createReq); err != nil {
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err), r.URL)
-		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
-	}
-
-	if err := createReq.Validate(); err != nil {
-		// Since this validation would happen client side as well, we only send
-		// a generic error message here.
-		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument), r.URL)
+	ctx, cred, opts, createReq, targetUser, APIError := commonAddServiceAccount(r)
+	if APIError.Code != "" {
+		writeErrorResponseJSON(ctx, w, APIError, r.URL)
 		return
 	}
 
 	var (
-		targetUser   string
 		targetGroups []string
+		err          error
 	)
 
-	// If the request did not set a TargetUser, the service account is
-	// created for the request sender.
-	targetUser = createReq.TargetUser
-	if targetUser == "" {
-		targetUser = cred.AccessKey
-	}
-
-	description := createReq.Description
-	if description == "" {
-		description = createReq.Comment
-	}
-	opts := newServiceAccountOpts{
-		accessKey:   createReq.AccessKey,
-		secretKey:   createReq.SecretKey,
-		name:        createReq.Name,
-		description: description,
-		expiration:  createReq.Expiration,
-		claims:      make(map[string]interface{}),
-	}
-
 	// Find the user for the request sender (as it may be sent via a service
 	// account or STS account):
 	requestorUser := cred.AccessKey
@@ -706,23 +659,6 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 	// that targetUser is a real user (i.e. not derived
 	// credentials).
 	if isSvcAccForRequestor {
-		// Check if adding service account is explicitly denied.
-		//
-		// This allows turning off service accounts for request sender,
-		// if there is no deny statement this call is implicitly enabled.
-		if !globalIAMSys.IsAllowed(policy.Args{
-			AccountName:     requestorUser,
-			Groups:          requestorGroups,
-			Action:          policy.CreateServiceAccountAdminAction,
-			ConditionValues: getConditionValues(r, "", cred),
-			IsOwner:         owner,
-			Claims:          cred.Claims,
-			DenyOnly:        true,
-		}) {
-			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
-			return
-		}
-
 		if requestorIsDerivedCredential {
 			if requestorParentUser == "" {
 				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx,
@@ -741,32 +677,16 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 			}
 			opts.claims[k] = v
 		}
-	} else {
-		// Need permission if we are creating a service account for a
-		// user <> to the request sender
-		if !globalIAMSys.IsAllowed(policy.Args{
-			AccountName:     requestorUser,
-			Groups:          requestorGroups,
-			Action:          policy.CreateServiceAccountAdminAction,
-			ConditionValues: getConditionValues(r, "", cred),
-			IsOwner:         owner,
-			Claims:          cred.Claims,
-		}) {
-			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
-			return
-		}
-
+	} else if globalIAMSys.LDAPConfig.Enabled() {
 		// In case of LDAP we need to resolve the targetUser to a DN and
 		// query their groups:
-		if globalIAMSys.LDAPConfig.Enabled() {
-			opts.claims[ldapUserN] = targetUser // simple username
-			targetUser, targetGroups, err = globalIAMSys.LDAPConfig.LookupUserDN(targetUser)
-			if err != nil {
-				writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-				return
-			}
-			opts.claims[ldapUser] = targetUser // username DN
+		opts.claims[ldapUserN] = targetUser // simple username
+		targetUser, targetGroups, err = globalIAMSys.LDAPConfig.LookupUserDN(targetUser)
+		if err != nil {
+			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+			return
 		}
+		opts.claims[ldapUser] = targetUser // username DN
 
 		// NOTE: if not using LDAP, then internal IDP or open ID is
 		// being used - in the former, group info is enforced when
@@ -774,16 +694,6 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 		// latter, a group notion is not supported.
 	}
 
-	var sp *policy.Policy
-	if len(createReq.Policy) > 0 {
-		sp, err = policy.ParseConfig(bytes.NewReader(createReq.Policy))
-		if err != nil {
-			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-			return
-		}
-	}
-
-	opts.sessionPolicy = sp
 	newCred, updatedAt, err := globalIAMSys.NewServiceAccount(ctx, targetUser, targetGroups, opts)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -804,7 +714,7 @@ func (a adminAPIHandlers) AddServiceAccount(w http.ResponseWriter, r *http.Reque
 		return
 	}
 
-	encryptedData, err := madmin.EncryptData(password, data)
+	encryptedData, err := madmin.EncryptData(cred.SecretKey, data)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -865,25 +775,6 @@ func (a adminAPIHandlers) UpdateServiceAccount(w http.ResponseWriter, r *http.Re
 		return
 	}
 
-	if !globalIAMSys.IsAllowed(policy.Args{
-		AccountName:     cred.AccessKey,
-		Groups:          cred.Groups,
-		Action:          policy.UpdateServiceAccountAdminAction,
-		ConditionValues: getConditionValues(r, "", cred),
-		IsOwner:         owner,
-		Claims:          cred.Claims,
-	}) {
-		requestUser := cred.AccessKey
-		if cred.ParentUser != "" {
-			requestUser = cred.ParentUser
-		}
-
-		if requestUser != svcAccount.ParentUser {
-			writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
-			return
-		}
-	}
-
 	password := cred.SecretKey
 	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
 	if err != nil {
@@ -904,6 +795,31 @@ func (a adminAPIHandlers) UpdateServiceAccount(w http.ResponseWriter, r *http.Re
 		return
 	}
 
+	condValues := getConditionValues(r, "", cred)
+	addExpirationToCondValues(updateReq.NewExpiration, condValues)
+
+	// Permission checks:
+	//
+	// 1. Any type of account (i.e. access keys (previously/still called service
+	// accounts), STS accounts, internal IDP accounts, etc) with the
+	// policy.UpdateServiceAccountAdminAction permission can update any service
+	// account.
+	//
+	// 2. We would like to let a user update their own access keys, however it
+	// is currently blocked pending a re-design. Users are still able to delete
+	// and re-create them.
+	if !globalIAMSys.IsAllowed(policy.Args{
+		AccountName:     cred.AccessKey,
+		Groups:          cred.Groups,
+		Action:          policy.UpdateServiceAccountAdminAction,
+		ConditionValues: condValues,
+		IsOwner:         owner,
+		Claims:          cred.Claims,
+	}) {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAccessDenied), r.URL)
+		return
+	}
+
 	var sp *policy.Policy
 	if len(updateReq.NewPolicy) > 0 {
 		sp, err = policy.ParseConfig(bytes.NewReader(updateReq.NewPolicy))
@@ -911,6 +827,9 @@ func (a adminAPIHandlers) UpdateServiceAccount(w http.ResponseWriter, r *http.Re
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
 		}
+		if sp.Version == "" && len(sp.Statements) == 0 {
+			sp = nil
+		}
 	}
 	opts := updateServiceAccountOpts{
 		secretKey:     updateReq.NewSecretKey,
@@ -996,12 +915,15 @@ func (a adminAPIHandlers) InfoServiceAccount(w http.ResponseWriter, r *http.Requ
 		}
 	}
 
+	// if session policy is nil or empty, then it is implied policy
+	impliedPolicy := sessionPolicy == nil || (sessionPolicy.Version == "" && len(sessionPolicy.Statements) == 0)
+
 	var svcAccountPolicy policy.Policy
 
-	if sessionPolicy != nil {
+	if !impliedPolicy {
 		svcAccountPolicy = *sessionPolicy
 	} else {
-		policiesNames, err := globalIAMSys.PolicyDBGet(svcAccount.ParentUser, false)
+		policiesNames, err := globalIAMSys.PolicyDBGet(svcAccount.ParentUser, svcAccount.Groups...)
 		if err != nil {
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 			return
@@ -1025,7 +947,7 @@ func (a adminAPIHandlers) InfoServiceAccount(w http.ResponseWriter, r *http.Requ
 		Name:          svcAccount.Name,
 		Description:   svcAccount.Description,
 		AccountStatus: svcAccount.Status,
-		ImpliedPolicy: sessionPolicy == nil,
+		ImpliedPolicy: impliedPolicy,
 		Policy:        string(policyJSON),
 		Expiration:    expiration,
 	}
@@ -1147,7 +1069,11 @@ func (a adminAPIHandlers) DeleteServiceAccount(w http.ResponseWriter, r *http.Re
 
 	// We do not care if service account is readable or not at this point,
 	// since this is a delete call we shall allow it to be deleted if possible.
-	svcAccount, _, _ := globalIAMSys.GetServiceAccount(ctx, serviceAccount)
+	svcAccount, _, err := globalIAMSys.GetServiceAccount(ctx, serviceAccount)
+	if errors.Is(err, errNoSuchServiceAccount) {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrAdminServiceAccountNotFound), r.URL)
+		return
+	}
 
 	adminPrivilege := globalIAMSys.IsAllowed(policy.Args{
 		AccountName:     cred.AccessKey,
@@ -1344,7 +1270,7 @@ func (a adminAPIHandlers) AccountInfoHandler(w http.ResponseWriter, r *http.Requ
 		effectivePolicy = globalIAMSys.GetCombinedPolicy(policySetFromClaims.ToSlice()...)
 
 	default:
-		policies, err := globalIAMSys.PolicyDBGet(accountName, false, cred.Groups...)
+		policies, err := globalIAMSys.PolicyDBGet(accountName, cred.Groups...)
 		if err != nil {
 			logger.LogIf(ctx, err)
 			writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -2494,3 +2420,95 @@ func (a adminAPIHandlers) ImportIAM(w http.ResponseWriter, r *http.Request) {
 		}
 	}
 }
+
+func addExpirationToCondValues(exp *time.Time, condValues map[string][]string) {
+	if exp == nil {
+		return
+	}
+	condValues["DurationSeconds"] = []string{strconv.FormatInt(int64(exp.Sub(time.Now()).Seconds()), 10)}
+}
+
+func commonAddServiceAccount(r *http.Request) (context.Context, auth.Credentials, newServiceAccountOpts, madmin.AddServiceAccountReq, string, APIError) {
+	ctx := r.Context()
+
+	// Get current object layer instance.
+	objectAPI := newObjectLayerFn()
+	if objectAPI == nil || globalNotificationSys == nil {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErr(ErrServerNotInitialized)
+	}
+
+	cred, owner, s3Err := validateAdminSignature(ctx, r, "")
+	if s3Err != ErrNone {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErr(s3Err)
+	}
+
+	password := cred.SecretKey
+	reqBytes, err := madmin.DecryptData(password, io.LimitReader(r.Body, r.ContentLength))
+	if err != nil {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err)
+	}
+
+	var createReq madmin.AddServiceAccountReq
+	if err = json.Unmarshal(reqBytes, &createReq); err != nil {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErrWithErr(ErrAdminConfigBadJSON, err)
+	}
+
+	// service account access key cannot have space characters beginning and end of the string.
+	if hasSpaceBE(createReq.AccessKey) {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument)
+	}
+
+	if err := createReq.Validate(); err != nil {
+		// Since this validation would happen client side as well, we only send
+		// a generic error message here.
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErr(ErrAdminResourceInvalidArgument)
+	}
+	// If the request did not set a TargetUser, the service account is
+	// created for the request sender.
+	targetUser := createReq.TargetUser
+	if targetUser == "" {
+		targetUser = cred.AccessKey
+	}
+
+	description := createReq.Description
+	if description == "" {
+		description = createReq.Comment
+	}
+	opts := newServiceAccountOpts{
+		accessKey:   createReq.AccessKey,
+		secretKey:   createReq.SecretKey,
+		name:        createReq.Name,
+		description: description,
+		expiration:  createReq.Expiration,
+		claims:      make(map[string]interface{}),
+	}
+
+	condValues := getConditionValues(r, "", cred)
+	addExpirationToCondValues(createReq.Expiration, condValues)
+
+	// Check if action is allowed if creating access key for another user
+	// Check if action is explicitly denied if for self
+	if !globalIAMSys.IsAllowed(policy.Args{
+		AccountName:     cred.AccessKey,
+		Groups:          cred.Groups,
+		Action:          policy.CreateServiceAccountAdminAction,
+		ConditionValues: condValues,
+		IsOwner:         owner,
+		Claims:          cred.Claims,
+		DenyOnly:        (targetUser == cred.AccessKey || targetUser == cred.ParentUser),
+	}) {
+		return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", errorCodes.ToAPIErr(ErrAccessDenied)
+	}
+
+	var sp *policy.Policy
+	if len(createReq.Policy) > 0 {
+		sp, err = policy.ParseConfig(bytes.NewReader(createReq.Policy))
+		if err != nil {
+			return ctx, auth.Credentials{}, newServiceAccountOpts{}, madmin.AddServiceAccountReq{}, "", toAdminAPIErr(ctx, err)
+		}
+	}
+
+	opts.sessionPolicy = sp
+
+	return ctx, cred, opts, createReq, targetUser, APIError{}
+}

cmd/admin-handlers-users_test.go

@@ -206,7 +206,9 @@ func TestIAMInternalIDPServerSuite(t *testing.T) {
 				suite.TestCannedPolicies(c)
 				suite.TestGroupAddRemove(c)
 				suite.TestServiceAccountOpsByAdmin(c)
+				suite.TestServiceAccountPrivilegeEscalationBug(c)
 				suite.TestServiceAccountOpsByUser(c)
+				suite.TestServiceAccountDurationSecondsCondition(c)
 				suite.TestAddServiceAccountPerms(c)
 				suite.TearDownSuite(c)
 			},
@@ -896,14 +898,6 @@ func (s *TestSuiteIAM) TestServiceAccountOpsByUser(c *check) {
 	// 3. Check S3 access
 	c.assertSvcAccS3Access(ctx, s, cr, bucket)
 
-	// 4. Check that svc account can restrict the policy, and that the
-	// session policy can be updated.
-	c.assertSvcAccSessionPolicyUpdate(ctx, s, userAdmClient, accessKey, 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)
 
@@ -911,6 +905,93 @@ func (s *TestSuiteIAM) TestServiceAccountOpsByUser(c *check) {
 	c.mustNotCreateSvcAccount(ctx, globalActiveCred.AccessKey, userAdmClient)
 }
 
+func (s *TestSuiteIAM) TestServiceAccountDurationSecondsCondition(c *check) {
+	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
+	defer cancel()
+
+	bucket := getRandomBucketName()
+	err := s.client.MakeBucket(ctx, bucket, minio.MakeBucketOptions{})
+	if err != nil {
+		c.Fatalf("bucket creat error: %v", err)
+	}
+
+	// Create policy, user and associate policy
+	policy := "mypolicy"
+	policyBytes := []byte(fmt.Sprintf(`{
+ "Version": "2012-10-17",
+ "Statement": [
+  {
+   "Effect": "Deny",
+   "Action": [
+     "admin:CreateServiceAccount",
+     "admin:UpdateServiceAccount"
+   ],
+   "Condition": {"NumericGreaterThan": {"svc:DurationSeconds": "3600"}}
+  },
+  {
+   "Effect": "Allow",
+   "Action": [
+    "s3:PutObject",
+    "s3:GetObject",
+    "s3:ListBucket"
+   ],
+   "Resource": [
+    "arn:aws:s3:::%s/*"
+   ]
+  }
+ ]
+}`, bucket))
+	err = s.adm.AddCannedPolicy(ctx, policy, policyBytes)
+	if err != nil {
+		c.Fatalf("policy add error: %v", err)
+	}
+
+	accessKey, secretKey := mustGenerateCredentials(c)
+	err = s.adm.SetUser(ctx, accessKey, secretKey, madmin.AccountEnabled)
+	if err != nil {
+		c.Fatalf("Unable to set user: %v", err)
+	}
+
+	err = s.adm.SetPolicy(ctx, policy, accessKey, false)
+	if err != nil {
+		c.Fatalf("Unable to set policy: %v", err)
+	}
+
+	// Create an madmin client with user creds
+	userAdmClient, err := madmin.NewWithOptions(s.endpoint, &madmin.Options{
+		Creds:  credentials.NewStaticV4(accessKey, secretKey, ""),
+		Secure: s.secure,
+	})
+	if err != nil {
+		c.Fatalf("Err creating user admin client: %v", err)
+	}
+	userAdmClient.SetCustomTransport(s.TestSuiteCommon.client.Transport)
+
+	distantExpiration := time.Now().Add(30 * time.Minute)
+	cr, err := userAdmClient.AddServiceAccount(ctx, madmin.AddServiceAccountReq{
+		TargetUser: accessKey,
+		AccessKey:  "svc-accesskey",
+		SecretKey:  "svc-secretkey",
+		Expiration: &distantExpiration,
+	})
+	if err != nil {
+		c.Fatalf("Unable to create svc acc: %v", err)
+	}
+
+	c.assertSvcAccS3Access(ctx, s, cr, bucket)
+
+	closeExpiration := time.Now().Add(2 * time.Hour)
+	_, err = userAdmClient.AddServiceAccount(ctx, madmin.AddServiceAccountReq{
+		TargetUser: accessKey,
+		AccessKey:  "svc-accesskey",
+		SecretKey:  "svc-secretkey",
+		Expiration: &closeExpiration,
+	})
+	if err == nil {
+		c.Fatalf("Creating a svc acc with distant expiration should fail")
+	}
+}
+
 func (s *TestSuiteIAM) TestServiceAccountOpsByAdmin(c *check) {
 	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
 	defer cancel()
@@ -979,6 +1060,95 @@ func (s *TestSuiteIAM) TestServiceAccountOpsByAdmin(c *check) {
 	c.assertSvcAccDeletion(ctx, s, s.adm, accessKey, bucket)
 }
 
+func (s *TestSuiteIAM) TestServiceAccountPrivilegeEscalationBug(c *check) {
+	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
+	defer cancel()
+
+	err := s.client.MakeBucket(ctx, "public", minio.MakeBucketOptions{})
+	if err != nil {
+		c.Fatalf("bucket creat error: %v", err)
+	}
+
+	err = s.client.MakeBucket(ctx, "private", minio.MakeBucketOptions{})
+	if err != nil {
+		c.Fatalf("bucket creat error: %v", err)
+	}
+
+	pubPolicyBytes := []byte(`{
+ "Version": "2012-10-17",
+ "Statement": [
+  {
+   "Effect": "Allow",
+   "Action": [
+    "s3:*"
+   ],
+   "Resource": [
+    "arn:aws:s3:::public",
+    "arn:aws:s3:::public/*"
+   ]
+  }
+ ]
+}`)
+
+	fullS3PolicyBytes := []byte(`{
+  "Version": "2012-10-17",
+  "Statement": [
+    {
+      "Effect": "Allow",
+      "Action": [
+        "s3:*"
+      ],
+      "Resource": [
+        "arn:aws:s3:::*"
+      ]
+    }
+  ]
+}
+`)
+
+	// Create a service account for the root user.
+	cr, err := s.adm.AddServiceAccount(ctx, madmin.AddServiceAccountReq{
+		TargetUser: globalActiveCred.AccessKey,
+		Policy:     pubPolicyBytes,
+	})
+	if err != nil {
+		c.Fatalf("admin should be able to create service account for themselves %s", err)
+	}
+
+	svcClient := s.getUserClient(c, cr.AccessKey, cr.SecretKey, "")
+
+	// Check that the service account can access the public bucket.
+	buckets, err := svcClient.ListBuckets(ctx)
+	if err != nil {
+		c.Fatalf("err fetching buckets %s", err)
+	}
+	if len(buckets) != 1 || buckets[0].Name != "public" {
+		c.Fatalf("service account should only have access to public bucket")
+	}
+
+	// Create an madmin client with the service account creds.
+	svcAdmClient, err := madmin.NewWithOptions(s.endpoint, &madmin.Options{
+		Creds:  credentials.NewStaticV4(cr.AccessKey, cr.SecretKey, ""),
+		Secure: s.secure,
+	})
+	if err != nil {
+		c.Fatalf("Err creating svcacct admin client: %v", err)
+	}
+	svcAdmClient.SetCustomTransport(s.TestSuiteCommon.client.Transport)
+
+	// Attempt to update the policy on the service account.
+	err = svcAdmClient.UpdateServiceAccount(ctx, cr.AccessKey,
+		madmin.UpdateServiceAccountReq{
+			NewPolicy: fullS3PolicyBytes,
+		})
+
+	if err == nil {
+		c.Fatalf("service account should not be able to update policy on itself")
+	} else if !strings.Contains(err.Error(), "Access Denied") {
+		c.Fatalf("unexpected error: %v", err)
+	}
+}
+
 func (s *TestSuiteIAM) SetUpAccMgmtPlugin(c *check) {
 	ctx, cancel := context.WithTimeout(context.Background(), testDefaultTimeout)
 	defer cancel()

cmd/admin-handlers.go

@@ -50,8 +50,10 @@ import (
 	"github.com/minio/madmin-go/v3/estream"
 	"github.com/minio/minio-go/v7/pkg/set"
 	"github.com/minio/minio/internal/dsync"
+	"github.com/minio/minio/internal/grid"
 	"github.com/minio/minio/internal/handlers"
 	xhttp "github.com/minio/minio/internal/http"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/kms"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/mux"
@@ -59,11 +61,14 @@ import (
 	xnet "github.com/minio/pkg/v2/net"
 	"github.com/minio/pkg/v2/policy"
 	"github.com/secure-io/sio-go"
+	"github.com/zeebo/xxh3"
 )
 
 const (
 	maxEConfigJSONSize        = 262272
 	kubernetesVersionEndpoint = "https://kubernetes.default.svc/version"
+	anonymizeParam            = "anonymize"
+	anonymizeStrict           = "strict"
 )
 
 // Only valid query params for mgmt admin APIs.
@@ -75,6 +80,213 @@ const (
 	mgmtForceStop   = "forceStop"
 )
 
+// ServerUpdateV2Handler - POST /minio/admin/v3/update?updateURL={updateURL}&type=2
+// ----------
+// updates all minio servers and restarts them gracefully.
+func (a adminAPIHandlers) ServerUpdateV2Handler(w http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	objectAPI, _ := validateAdminReq(ctx, w, r, policy.ServerUpdateAdminAction)
+	if objectAPI == nil {
+		return
+	}
+
+	if globalInplaceUpdateDisabled || currentReleaseTime.IsZero() {
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrMethodNotAllowed), r.URL)
+		return
+	}
+
+	vars := mux.Vars(r)
+	updateURL := vars["updateURL"]
+	dryRun := r.Form.Get("dry-run") == "true"
+
+	mode := getMinioMode()
+	if updateURL == "" {
+		updateURL = minioReleaseInfoURL
+		if runtime.GOOS == globalWindowsOSName {
+			updateURL = minioReleaseWindowsInfoURL
+		}
+	}
+
+	u, err := url.Parse(updateURL)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	content, err := downloadReleaseURL(u, updateTimeout, mode)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	sha256Sum, lrTime, releaseInfo, err := parseReleaseData(content)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	u.Path = path.Dir(u.Path) + SlashSeparator + releaseInfo
+	// Download Binary Once
+	binC, bin, err := downloadBinary(u, mode)
+	if err != nil {
+		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	updateStatus := madmin.ServerUpdateStatusV2{DryRun: dryRun}
+	peerResults := make(map[string]madmin.ServerPeerUpdateStatus)
+
+	local := globalLocalNodeName
+	if local == "" {
+		local = "127.0.0.1"
+	}
+
+	failedClients := make(map[int]struct{})
+
+	if globalIsDistErasure {
+		// Push binary to other servers
+		for idx, nerr := range globalNotificationSys.VerifyBinary(ctx, u, sha256Sum, releaseInfo, binC) {
+			if nerr.Err != nil {
+				peerResults[nerr.Host.String()] = madmin.ServerPeerUpdateStatus{
+					Host:           nerr.Host.String(),
+					Err:            nerr.Err.Error(),
+					CurrentVersion: Version,
+				}
+				failedClients[idx] = struct{}{}
+			} else {
+				peerResults[nerr.Host.String()] = madmin.ServerPeerUpdateStatus{
+					Host:           nerr.Host.String(),
+					CurrentVersion: Version,
+					UpdatedVersion: lrTime.Format(minioReleaseTagTimeLayout),
+				}
+			}
+		}
+	}
+
+	if lrTime.Sub(currentReleaseTime) > 0 {
+		if err = verifyBinary(u, sha256Sum, releaseInfo, mode, bytes.NewReader(bin)); err != nil {
+			peerResults[local] = madmin.ServerPeerUpdateStatus{
+				Host:           local,
+				Err:            err.Error(),
+				CurrentVersion: Version,
+			}
+		} else {
+			peerResults[local] = madmin.ServerPeerUpdateStatus{
+				Host:           local,
+				CurrentVersion: Version,
+				UpdatedVersion: lrTime.Format(minioReleaseTagTimeLayout),
+			}
+		}
+	} else {
+		peerResults[local] = madmin.ServerPeerUpdateStatus{
+			Host:           local,
+			Err:            fmt.Sprintf("server is already running the latest version: %s", Version),
+			CurrentVersion: Version,
+		}
+	}
+
+	if !dryRun {
+		if globalIsDistErasure {
+			ng := WithNPeers(len(globalNotificationSys.peerClients))
+			for idx, client := range globalNotificationSys.peerClients {
+				_, ok := failedClients[idx]
+				if ok {
+					continue
+				}
+				client := client
+				ng.Go(ctx, func() error {
+					return client.CommitBinary(ctx)
+				}, idx, *client.host)
+			}
+
+			for _, nerr := range ng.Wait() {
+				if nerr.Err != nil {
+					prs, ok := peerResults[nerr.Host.String()]
+					if ok {
+						prs.Err = nerr.Err.Error()
+						peerResults[nerr.Host.String()] = prs
+					} else {
+						peerResults[nerr.Host.String()] = madmin.ServerPeerUpdateStatus{
+							Host:           nerr.Host.String(),
+							Err:            nerr.Err.Error(),
+							CurrentVersion: Version,
+							UpdatedVersion: lrTime.Format(minioReleaseTagTimeLayout),
+						}
+					}
+				}
+			}
+		}
+		prs := peerResults[local]
+		if prs.Err == "" {
+			if err = commitBinary(); err != nil {
+				prs.Err = err.Error()
+			}
+			peerResults[local] = prs
+		}
+	}
+
+	prs, ok := peerResults[local]
+	if ok {
+		prs.WaitingDrives = waitingDrivesNode()
+		peerResults[local] = prs
+	}
+
+	if globalIsDistErasure {
+		// Notify all other MinIO peers signal service.
+		ng := WithNPeers(len(globalNotificationSys.peerClients))
+		for idx, client := range globalNotificationSys.peerClients {
+			_, ok := failedClients[idx]
+			if ok {
+				continue
+			}
+			client := client
+			ng.Go(ctx, func() error {
+				prs, ok := peerResults[client.String()]
+				if ok && prs.CurrentVersion != prs.UpdatedVersion && prs.UpdatedVersion != "" {
+					return client.SignalService(serviceRestart, "", dryRun)
+				}
+				return nil
+			}, idx, *client.host)
+		}
+
+		for _, nerr := range ng.Wait() {
+			if nerr.Err != nil {
+				waitingDrives := map[string]madmin.DiskMetrics{}
+				jerr := json.Unmarshal([]byte(nerr.Err.Error()), &waitingDrives)
+				if jerr == nil {
+					prs, ok := peerResults[nerr.Host.String()]
+					if ok {
+						prs.WaitingDrives = waitingDrives
+						peerResults[nerr.Host.String()] = prs
+					}
+					continue
+				}
+			}
+		}
+	}
+
+	for _, pr := range peerResults {
+		updateStatus.Results = append(updateStatus.Results, pr)
+	}
+
+	// Marshal API response
+	jsonBytes, err := json.Marshal(updateStatus)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	writeSuccessResponseJSON(w, jsonBytes)
+
+	if !dryRun {
+		if lrTime.Sub(currentReleaseTime) > 0 {
+			globalServiceSignalCh <- serviceRestart
+		}
+	}
+}
+
 // ServerUpdateHandler - POST /minio/admin/v3/update?updateURL={updateURL}
 // ----------
 // updates all minio servers and restarts them gracefully.
@@ -86,7 +298,7 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	if globalInplaceUpdateDisabled {
+	if globalInplaceUpdateDisabled || currentReleaseTime.IsZero() {
 		// if MINIO_UPDATE=off - inplace update is disabled, mostly in containers.
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrMethodNotAllowed), r.URL)
 		return
@@ -120,14 +332,7 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
-	u.Path = path.Dir(u.Path) + SlashSeparator + releaseInfo
-	crTime, err := GetCurrentReleaseTime()
-	if err != nil {
-		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
-		return
-	}
-
-	if lrTime.Sub(crTime) <= 0 {
+	if lrTime.Sub(currentReleaseTime) <= 0 {
 		updateStatus := madmin.ServerUpdateStatus{
 			CurrentVersion: Version,
 			UpdatedVersion: Version,
@@ -144,8 +349,10 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		return
 	}
 
+	u.Path = path.Dir(u.Path) + SlashSeparator + releaseInfo
+
 	// Download Binary Once
-	reader, err := downloadBinary(u, mode)
+	binC, bin, err := downloadBinary(u, mode)
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -153,7 +360,7 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 	}
 
 	// Push binary to other servers
-	for _, nerr := range globalNotificationSys.VerifyBinary(ctx, u, sha256Sum, releaseInfo, reader) {
+	for _, nerr := range globalNotificationSys.VerifyBinary(ctx, u, sha256Sum, releaseInfo, binC) {
 		if nerr.Err != nil {
 			err := AdminError{
 				Code:       AdminUpdateApplyFailure,
@@ -167,7 +374,7 @@ func (a adminAPIHandlers) ServerUpdateHandler(w http.ResponseWriter, r *http.Req
 		}
 	}
 
-	err = verifyBinary(u, sha256Sum, releaseInfo, mode, reader)
+	err = verifyBinary(u, sha256Sum, releaseInfo, mode, bytes.NewReader(bin))
 	if err != nil {
 		logger.LogIf(ctx, fmt.Errorf("server update failed with %w", err))
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
@@ -244,7 +451,7 @@ func (a adminAPIHandlers) ServiceHandler(w http.ResponseWriter, r *http.Request)
 	case madmin.ServiceActionUnfreeze:
 		serviceSig = serviceUnFreeze
 	default:
-		logger.LogIf(ctx, fmt.Errorf("Unrecognized service action %s requested", action), logger.Application)
+		logger.LogIf(ctx, fmt.Errorf("Unrecognized service action %s requested", action), logger.ErrorKind)
 		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrMalformedPOSTRequest), r.URL)
 		return
 	}
@@ -283,6 +490,131 @@ func (a adminAPIHandlers) ServiceHandler(w http.ResponseWriter, r *http.Request)
 	}
 }
 
+type servicePeerResult struct {
+	Host          string                        `json:"host"`
+	Err           string                        `json:"err,omitempty"`
+	WaitingDrives map[string]madmin.DiskMetrics `json:"waitingDrives,omitempty"`
+}
+
+type serviceResult struct {
+	Action  madmin.ServiceAction `json:"action"`
+	DryRun  bool                 `json:"dryRun"`
+	Results []servicePeerResult  `json:"results,omitempty"`
+}
+
+// ServiceV2Handler - POST /minio/admin/v3/service?action={action}&type=2
+// ----------
+// Supports following actions:
+// - restart (restarts all the MinIO instances in a setup)
+// - stop (stops all the MinIO instances in a setup)
+// - freeze (freezes all incoming S3 API calls)
+// - unfreeze (unfreezes previously frozen S3 API calls)
+//
+// This newer API now returns back status per remote peer and local regarding
+// if a "restart/stop" was successful or not. Service signal now supports
+// a dry-run that helps skip the nodes that may have hung drives. By default
+// restart/stop will ignore the servers that are hung on drives. You can use
+// 'force' param to force restart even with hung drives if needed.
+func (a adminAPIHandlers) ServiceV2Handler(w http.ResponseWriter, r *http.Request) {
+	ctx := r.Context()
+
+	vars := mux.Vars(r)
+	action := vars["action"]
+	dryRun := r.Form.Get("dry-run") == "true"
+
+	var serviceSig serviceSignal
+	act := madmin.ServiceAction(action)
+	switch act {
+	case madmin.ServiceActionRestart:
+		serviceSig = serviceRestart
+	case madmin.ServiceActionStop:
+		serviceSig = serviceStop
+	case madmin.ServiceActionFreeze:
+		serviceSig = serviceFreeze
+	case madmin.ServiceActionUnfreeze:
+		serviceSig = serviceUnFreeze
+	default:
+		logger.LogIf(ctx, fmt.Errorf("Unrecognized service action %s requested", action), logger.ErrorKind)
+		writeErrorResponseJSON(ctx, w, errorCodes.ToAPIErr(ErrMalformedPOSTRequest), r.URL)
+		return
+	}
+
+	var objectAPI ObjectLayer
+	switch serviceSig {
+	case serviceRestart:
+		objectAPI, _ = validateAdminReq(ctx, w, r, policy.ServiceRestartAdminAction)
+	case serviceStop:
+		objectAPI, _ = validateAdminReq(ctx, w, r, policy.ServiceStopAdminAction)
+	case serviceFreeze, serviceUnFreeze:
+		objectAPI, _ = validateAdminReq(ctx, w, r, policy.ServiceFreezeAdminAction)
+	}
+	if objectAPI == nil {
+		return
+	}
+
+	// Notify all other MinIO peers signal service.
+	srvResult := serviceResult{Action: act, Results: []servicePeerResult{}}
+
+	process := act == madmin.ServiceActionRestart || act == madmin.ServiceActionStop
+	if process {
+		localhost := globalLocalNodeName
+		if globalLocalNodeName == "" {
+			localhost = "127.0.0.1"
+		}
+		waitingDrives := waitingDrivesNode()
+		srvResult.Results = append(srvResult.Results, servicePeerResult{
+			Host:          localhost,
+			WaitingDrives: waitingDrives,
+		})
+	}
+
+	if globalIsDistErasure {
+		for _, nerr := range globalNotificationSys.SignalServiceV2(serviceSig, dryRun) {
+			if nerr.Err != nil && process {
+				waitingDrives := map[string]madmin.DiskMetrics{}
+				jerr := json.Unmarshal([]byte(nerr.Err.Error()), &waitingDrives)
+				if jerr == nil {
+					srvResult.Results = append(srvResult.Results, servicePeerResult{
+						Host:          nerr.Host.String(),
+						WaitingDrives: waitingDrives,
+					})
+					continue
+				}
+			}
+			errStr := ""
+			if nerr.Err != nil {
+				errStr = nerr.Err.Error()
+			}
+			srvResult.Results = append(srvResult.Results, servicePeerResult{
+				Host: nerr.Host.String(),
+				Err:  errStr,
+			})
+		}
+	}
+
+	srvResult.DryRun = dryRun
+
+	buf, err := json.Marshal(srvResult)
+	if err != nil {
+		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
+		return
+	}
+
+	// Reply to the client before restarting, stopping MinIO server.
+	writeSuccessResponseJSON(w, buf)
+
+	switch serviceSig {
+	case serviceFreeze:
+		freezeServices()
+	case serviceUnFreeze:
+		unfreezeServices()
+	case serviceRestart, serviceStop:
+		if !dryRun {
+			globalServiceSignalCh <- serviceSig
+		}
+	}
+}
+
 // ServerProperties holds some server information such as, version, region
 // uptime, etc..
 type ServerProperties struct {
@@ -336,7 +668,7 @@ func (a adminAPIHandlers) StorageInfoHandler(w http.ResponseWriter, r *http.Requ
 		return
 	}
 
-	storageInfo := objectAPI.StorageInfo(ctx)
+	storageInfo := objectAPI.StorageInfo(ctx, true)
 
 	// Collect any disk healing.
 	healing, _ := getAggregatedBackgroundHealState(ctx, nil)
@@ -473,7 +805,7 @@ func (a adminAPIHandlers) MetricsHandler(w http.ResponseWriter, r *http.Request)
 	}
 }
 
-// DataUsageInfoHandler - GET /minio/admin/v3/datausage
+// DataUsageInfoHandler - GET /minio/admin/v3/datausage?capacity={true}
 // ----------
 // Get server/cluster data usage info
 func (a adminAPIHandlers) DataUsageInfoHandler(w http.ResponseWriter, r *http.Request) {
@@ -496,6 +828,16 @@ func (a adminAPIHandlers) DataUsageInfoHandler(w http.ResponseWriter, r *http.Re
 		return
 	}
 
+	// Get capacity info when asked.
+	if r.Form.Get("capacity") == "true" {
+		sinfo := objectAPI.StorageInfo(ctx, false)
+		dataUsageInfo.TotalCapacity = GetTotalUsableCapacity(sinfo.Disks, sinfo)
+		dataUsageInfo.TotalFreeCapacity = GetTotalUsableCapacityFree(sinfo.Disks, sinfo)
+		if dataUsageInfo.TotalCapacity > dataUsageInfo.TotalFreeCapacity {
+			dataUsageInfo.TotalUsedCapacity = dataUsageInfo.TotalCapacity - dataUsageInfo.TotalFreeCapacity
+		}
+	}
+
 	writeSuccessResponseJSON(w, dataUsageInfoJSON)
 }
 
@@ -754,11 +1096,8 @@ func (a adminAPIHandlers) ProfileHandler(w http.ResponseWriter, r *http.Request)
 			return
 		}
 	}
-	// read request body
-	io.CopyN(io.Discard, r.Body, 1)
 
 	globalProfilerMu.Lock()
-
 	if globalProfiler == nil {
 		globalProfiler = make(map[string]minioProfiler, 10)
 	}
@@ -900,7 +1239,7 @@ func extractHealInitParams(vars map[string]string, qParms url.Values, r io.Reade
 	if hip.clientToken == "" {
 		jerr := json.NewDecoder(r).Decode(&hip.hs)
 		if jerr != nil {
-			logger.LogIf(GlobalContext, jerr, logger.Application)
+			logger.LogIf(GlobalContext, jerr, logger.ErrorKind)
 			err = ErrRequestBodyParse
 			return
 		}
@@ -1001,7 +1340,7 @@ func (a adminAPIHandlers) HealHandler(w http.ResponseWriter, r *http.Request) {
 							Message:   hr.errBody,
 							Resource:  r.URL.Path,
 							RequestID: w.Header().Get(xhttp.AmzRequestID),
-							HostID:    globalDeploymentID,
+							HostID:    globalDeploymentID(),
 						})
 					}
 					if !started {
@@ -1056,7 +1395,7 @@ func (a adminAPIHandlers) HealHandler(w http.ResponseWriter, r *http.Request) {
 		return
 	}
 
-	respCh := make(chan healResp)
+	respCh := make(chan healResp, 1)
 	switch {
 	case hip.forceStop:
 		go func() {
@@ -1218,7 +1557,7 @@ func (a adminAPIHandlers) ClientDevNull(w http.ResponseWriter, r *http.Request)
 	totalRx := int64(0)
 	connectTime := time.Now()
 	for {
-		n, err := io.CopyN(io.Discard, r.Body, 128*humanize.KiByte)
+		n, err := io.CopyN(xioutil.Discard, r.Body, 128*humanize.KiByte)
 		if err != nil && err != io.EOF && err != io.ErrUnexpectedEOF {
 			// would mean the network is not stable. Logging here will help in debugging network issues.
 			if time.Since(connectTime) < (globalNetPerfMinDuration - time.Second) {
@@ -1296,6 +1635,7 @@ func (a adminAPIHandlers) ObjectSpeedTestHandler(w http.ResponseWriter, r *http.
 	customBucket := strings.TrimSpace(r.Form.Get(peerRESTBucket))
 	autotune := r.Form.Get("autotune") == "true"
 	noClear := r.Form.Get("noclear") == "true"
+	enableSha256 := r.Form.Get("enableSha256") == "true"
 
 	size, err := strconv.Atoi(sizeStr)
 	if err != nil {
@@ -1312,7 +1652,7 @@ func (a adminAPIHandlers) ObjectSpeedTestHandler(w http.ResponseWriter, r *http.
 		duration = time.Second * 10
 	}
 
-	storageInfo := objectAPI.StorageInfo(ctx)
+	storageInfo := objectAPI.StorageInfo(ctx, false)
 
 	sufficientCapacity, canAutotune, capacityErrMsg := validateObjPerfOptions(ctx, storageInfo, concurrent, size, autotune)
 	if !sufficientCapacity {
@@ -1365,6 +1705,7 @@ func (a adminAPIHandlers) ObjectSpeedTestHandler(w http.ResponseWriter, r *http.
 		autotune:         autotune,
 		storageClass:     storageClass,
 		bucketName:       customBucket,
+		enableSha256:     enableSha256,
 	})
 	var prevResult madmin.SpeedTestResult
 	for {
@@ -1400,13 +1741,21 @@ func (a adminAPIHandlers) ObjectSpeedTestHandler(w http.ResponseWriter, r *http.
 }
 
 func makeObjectPerfBucket(ctx context.Context, objectAPI ObjectLayer, bucketName string) (bucketExists bool, err error) {
-	if err = objectAPI.MakeBucket(ctx, bucketName, MakeBucketOptions{}); err != nil {
+	if err = objectAPI.MakeBucket(ctx, bucketName, MakeBucketOptions{VersioningEnabled: globalSiteReplicationSys.isEnabled()}); err != nil {
 		if _, ok := err.(BucketExists); !ok {
 			// Only BucketExists error can be ignored.
 			return false, err
 		}
 		bucketExists = true
 	}
+
+	if globalSiteReplicationSys.isEnabled() {
+		configData := []byte(`<VersioningConfiguration><Status>Enabled</Status><ExcludedPrefixes><Prefix>speedtest/*</Prefix></ExcludedPrefixes></VersioningConfiguration>`)
+		if _, err = globalBucketMetadataSys.Update(ctx, bucketName, bucketVersioningConfig, configData); err != nil {
+			return false, err
+		}
+	}
+
 	return bucketExists, nil
 }
 
@@ -1580,14 +1929,13 @@ func (a adminAPIHandlers) TraceHandler(w http.ResponseWriter, r *http.Request) {
 	setEventStreamHeaders(w)
 
 	// Trace Publisher and peer-trace-client uses nonblocking send and hence does not wait for slow receivers.
-	// Use buffered channel to take care of burst sends or slow w.Write()
-	traceCh := make(chan madmin.TraceInfo, 4000)
-
+	// Keep 100k buffered channel.
+	// If receiver cannot keep up with that we drop events.
+	traceCh := make(chan []byte, 100000)
 	peers, _ := newPeerRestClients(globalEndpoints)
-
-	err = globalTrace.Subscribe(traceOpts.TraceTypes(), traceCh, ctx.Done(), func(entry madmin.TraceInfo) bool {
+	err = globalTrace.SubscribeJSON(traceOpts.TraceTypes(), traceCh, ctx.Done(), func(entry madmin.TraceInfo) bool {
 		return shouldTrace(entry, traceOpts)
-	})
+	}, nil)
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -1602,19 +1950,19 @@ func (a adminAPIHandlers) TraceHandler(w http.ResponseWriter, r *http.Request) {
 		if peer == nil {
 			continue
 		}
-		peer.Trace(traceCh, ctx.Done(), traceOpts)
+		peer.Trace(ctx, traceCh, traceOpts)
 	}
 
-	keepAliveTicker := time.NewTicker(500 * time.Millisecond)
+	keepAliveTicker := time.NewTicker(time.Second)
 	defer keepAliveTicker.Stop()
 
-	enc := json.NewEncoder(w)
 	for {
 		select {
 		case entry := <-traceCh:
-			if err := enc.Encode(entry); err != nil {
+			if _, err := w.Write(entry); err != nil {
 				return
 			}
+			grid.PutByteBuffer(entry)
 			if len(traceCh) == 0 {
 				// Flush if nothing is queued
 				w.(http.Flusher).Flush()
@@ -1853,7 +2201,10 @@ func getPoolsInfo(ctx context.Context, allDisks []madmin.Disk) (map[int]map[int]
 		return nil, errServerNotInitialized
 	}
 
-	z, _ := objectAPI.(*erasureServerPools)
+	z, ok := objectAPI.(*erasureServerPools)
+	if !ok {
+		return nil, errServerNotInitialized
+	}
 
 	poolsInfo := make(map[int]map[int]madmin.ErasureSetInfo)
 	for _, d := range allDisks {
@@ -1884,7 +2235,7 @@ func getPoolsInfo(ctx context.Context, allDisks []madmin.Disk) (map[int]map[int]
 	return poolsInfo, nil
 }
 
-func getServerInfo(ctx context.Context, poolsInfoEnabled bool, r *http.Request) madmin.InfoMessage {
+func getServerInfo(ctx context.Context, pools, metrics bool, r *http.Request) madmin.InfoMessage {
 	ldap := madmin.LDAP{}
 	if globalIAMSys.LDAPConfig.Enabled() {
 		ldapConn, err := globalIAMSys.LDAPConfig.LDAP.Connect()
@@ -1900,13 +2251,13 @@ func getServerInfo(ctx context.Context, poolsInfoEnabled bool, r *http.Request)
 		}
 	}
 
-	log, audit := fetchLoggerInfo()
+	log, audit := fetchLoggerInfo(ctx)
 
 	// Get the notification target info
 	notifyTarget := fetchLambdaInfo()
 
-	local := getLocalServerProperty(globalEndpoints, r)
-	servers := globalNotificationSys.ServerInfo()
+	local := getLocalServerProperty(globalEndpoints, r, metrics)
+	servers := globalNotificationSys.ServerInfo(metrics)
 	servers = append(servers, local)
 
 	assignPoolNumbers(servers)
@@ -1960,7 +2311,7 @@ func getServerInfo(ctx context.Context, poolsInfoEnabled bool, r *http.Request)
 			DrivesPerSet:     backendInfo.DrivesPerSet,
 		}
 
-		if poolsInfoEnabled {
+		if pools {
 			poolsInfo, _ = getPoolsInfo(ctx, allDisks)
 		}
 	}
@@ -1980,7 +2331,7 @@ func getServerInfo(ctx context.Context, poolsInfoEnabled bool, r *http.Request)
 		Domain:        domain,
 		Region:        globalSite.Region,
 		SQSARN:        globalEventNotifier.GetARNList(false),
-		DeploymentID:  globalDeploymentID,
+		DeploymentID:  globalDeploymentID(),
 		Buckets:       buckets,
 		Objects:       objects,
 		Versions:      versions,
@@ -2006,7 +2357,7 @@ func getKubernetesInfo(dctx context.Context) madmin.KubernetesInfo {
 	}
 
 	client := &http.Client{
-		Transport: NewHTTPTransport(),
+		Transport: globalHealthChkTransport,
 		Timeout:   10 * time.Second,
 	}
 
@@ -2025,8 +2376,14 @@ func getKubernetesInfo(dctx context.Context) madmin.KubernetesInfo {
 
 func fetchHealthInfo(healthCtx context.Context, objectAPI ObjectLayer, query *url.Values, healthInfoCh chan madmin.HealthInfo, healthInfo madmin.HealthInfo) {
 	hostAnonymizer := createHostAnonymizer()
-	// anonAddr - Anonymizes hosts in given input string.
+
+	anonParam := query.Get(anonymizeParam)
+	// anonAddr - Anonymizes hosts in given input string
+	// (only if the anonymize param is set to srict).
 	anonAddr := func(addr string) string {
+		if anonParam != anonymizeStrict {
+			return addr
+		}
 		newAddr, found := hostAnonymizer[addr]
 		if found {
 			return newAddr
@@ -2087,6 +2444,20 @@ func fetchHealthInfo(healthCtx context.Context, objectAPI ObjectLayer, query *ur
 		}
 	}
 
+	getAndWriteNetInfo := func() {
+		if query.Get(string(madmin.HealthDataTypeSysNet)) == "true" {
+			localNetInfo := madmin.GetNetInfo(globalLocalNodeName, globalInternodeInterface)
+			healthInfo.Sys.NetInfo = append(healthInfo.Sys.NetInfo, localNetInfo)
+
+			peerNetInfos := globalNotificationSys.GetNetInfo(healthCtx)
+			for _, n := range peerNetInfos {
+				anonymizeAddr(&n)
+				healthInfo.Sys.NetInfo = append(healthInfo.Sys.NetInfo, n)
+			}
+			partialWrite(healthInfo)
+		}
+	}
+
 	getAndWriteOSInfo := func() {
 		if query.Get("sysosinfo") == "true" {
 			localOSInfo := madmin.GetOSInfo(healthCtx, globalLocalNodeName)
@@ -2187,6 +2558,10 @@ func fetchHealthInfo(healthCtx context.Context, objectAPI ObjectLayer, query *ur
 	}
 
 	anonymizeCmdLine := func(cmdLine string) string {
+		if anonParam != anonymizeStrict {
+			return cmdLine
+		}
+
 		if !globalIsDistErasure {
 			// FS mode - single server - hard code to `server1`
 			anonCmdLine := strings.ReplaceAll(cmdLine, globalLocalNodeName, "server1")
@@ -2302,12 +2677,13 @@ func fetchHealthInfo(healthCtx context.Context, objectAPI ObjectLayer, query *ur
 	}
 
 	go func() {
-		defer close(healthInfoCh)
+		defer xioutil.SafeClose(healthInfoCh)
 
 		partialWrite(healthInfo) // Write first message with only version and deployment id populated
 		getAndWritePlatformInfo()
 		getAndWriteCPUs()
 		getAndWritePartitions()
+		getAndWriteNetInfo()
 		getAndWriteOSInfo()
 		getAndWriteMemInfo()
 		getAndWriteProcInfo()
@@ -2317,7 +2693,7 @@ func fetchHealthInfo(healthCtx context.Context, objectAPI ObjectLayer, query *ur
 		getAndWriteSysConfig()
 
 		if query.Get("minioinfo") == "true" {
-			infoMessage := getServerInfo(healthCtx, false, nil)
+			infoMessage := getServerInfo(healthCtx, false, true, nil)
 			servers := make([]madmin.ServerInfo, 0, len(infoMessage.Servers))
 			for _, server := range infoMessage.Servers {
 				anonEndpoint := anonAddr(server.Endpoint)
@@ -2390,7 +2766,7 @@ func (a adminAPIHandlers) HealthInfoHandler(w http.ResponseWriter, r *http.Reque
 		Version:   madmin.HealthInfoVersion,
 		Minio: madmin.MinioHealthInfo{
 			Info: madmin.MinioInfo{
-				DeploymentID: globalDeploymentID,
+				DeploymentID: globalDeploymentID(),
 			},
 		},
 	}
@@ -2465,11 +2841,27 @@ func getTLSInfo() madmin.TLSInfo {
 
 	if globalIsTLS {
 		for _, c := range globalPublicCerts {
+			check := xxh3.Hash(c.RawIssuer)
+			check ^= xxh3.Hash(c.RawSubjectPublicKeyInfo)
+			// We XOR, so order doesn't matter.
+			for _, v := range c.DNSNames {
+				check ^= xxh3.HashString(v)
+			}
+			for _, v := range c.EmailAddresses {
+				check ^= xxh3.HashString(v)
+			}
+			for _, v := range c.IPAddresses {
+				check ^= xxh3.HashString(v.String())
+			}
+			for _, v := range c.URIs {
+				check ^= xxh3.HashString(v.String())
+			}
 			tlsInfo.Certs = append(tlsInfo.Certs, madmin.TLSCert{
 				PubKeyAlgo:    c.PublicKeyAlgorithm.String(),
 				SignatureAlgo: c.SignatureAlgorithm.String(),
 				NotBefore:     c.NotBefore,
 				NotAfter:      c.NotAfter,
+				Checksum:      strconv.FormatUint(check, 16),
 			})
 		}
 	}
@@ -2489,8 +2881,10 @@ func (a adminAPIHandlers) ServerInfoHandler(w http.ResponseWriter, r *http.Reque
 		return
 	}
 
+	metrics := r.Form.Get(peerRESTMetrics) == "true"
+
 	// Marshal API response
-	jsonBytes, err := json.Marshal(getServerInfo(ctx, true, r))
+	jsonBytes, err := json.Marshal(getServerInfo(ctx, true, metrics, r))
 	if err != nil {
 		writeErrorResponseJSON(ctx, w, toAdminAPIErr(ctx, err), r.URL)
 		return
@@ -2520,7 +2914,7 @@ func assignPoolNumbers(servers []madmin.ServerProperties) {
 func fetchLambdaInfo() []map[string][]madmin.TargetIDStatus {
 	lambdaMap := make(map[string][]madmin.TargetIDStatus)
 
-	for _, tgt := range globalNotifyTargetList.Targets() {
+	for _, tgt := range globalEventNotifier.Targets() {
 		targetIDStatus := make(map[string]madmin.Status)
 		active, _ := tgt.IsActive()
 		targetID := tgt.ID()
@@ -2608,19 +3002,26 @@ func fetchKMSStatusV2(ctx context.Context) []madmin.KMS {
 	return stats
 }
 
+func targetStatus(ctx context.Context, h logger.Target) madmin.Status {
+	if h.IsOnline(ctx) {
+		return madmin.Status{Status: string(madmin.ItemOnline)}
+	}
+	return madmin.Status{Status: string(madmin.ItemOffline)}
+}
+
 // fetchLoggerDetails return log info
-func fetchLoggerInfo() ([]madmin.Logger, []madmin.Audit) {
+func fetchLoggerInfo(ctx context.Context) ([]madmin.Logger, []madmin.Audit) {
 	var loggerInfo []madmin.Logger
 	var auditloggerInfo []madmin.Audit
 	for _, tgt := range logger.SystemTargets() {
 		if tgt.Endpoint() != "" {
-			loggerInfo = append(loggerInfo, madmin.Logger{tgt.String(): logger.TargetStatus(GlobalContext, tgt)})
+			loggerInfo = append(loggerInfo, madmin.Logger{tgt.String(): targetStatus(ctx, tgt)})
 		}
 	}
 
 	for _, tgt := range logger.AuditTargets() {
 		if tgt.Endpoint() != "" {
-			auditloggerInfo = append(auditloggerInfo, madmin.Audit{tgt.String(): logger.TargetStatus(GlobalContext, tgt)})
+			auditloggerInfo = append(auditloggerInfo, madmin.Audit{tgt.String(): targetStatus(ctx, tgt)})
 		}
 	}
 
@@ -2667,14 +3068,14 @@ func getClusterMetaInfo(ctx context.Context) []byte {
 	resultCh := make(chan madmin.ClusterRegistrationInfo)
 
 	go func() {
-		defer close(resultCh)
+		defer xioutil.SafeClose(resultCh)
 
 		ci := madmin.ClusterRegistrationInfo{}
 		ci.Info.NoOfServerPools = len(globalEndpoints)
-		ci.Info.NoOfServers = len(globalEndpoints.Hostnames())
+		ci.Info.NoOfServers = totalNodeCount()
 		ci.Info.MinioVersion = Version
 
-		si := objectAPI.StorageInfo(ctx)
+		si := objectAPI.StorageInfo(ctx, false)
 
 		ci.Info.NoOfDrives = len(si.Disks)
 		for _, disk := range si.Disks {
@@ -2688,7 +3089,7 @@ func getClusterMetaInfo(ctx context.Context) []byte {
 		ci.Info.NoOfBuckets = dataUsageInfo.BucketsCount
 		ci.Info.NoOfObjects = dataUsageInfo.ObjectsTotalCount
 
-		ci.DeploymentID = globalDeploymentID
+		ci.DeploymentID = globalDeploymentID()
 		ci.ClusterName = fmt.Sprintf("%d-servers-%d-disks-%s", ci.Info.NoOfServers, ci.Info.NoOfDrives, ci.Info.MinioVersion)
 
 		select {
@@ -2911,7 +3312,7 @@ func (a adminAPIHandlers) InspectDataHandler(w http.ResponseWriter, r *http.Requ
 	}
 
 	// save the format.json as part of inspect by default
-	if volume != minioMetaBucket && file != formatConfigFile {
+	if !(volume == minioMetaBucket && file == formatConfigFile) {
 		err = o.GetRawData(ctx, minioMetaBucket, formatConfigFile, rawDataFn)
 	}
 	if !errors.Is(err, errFileNotFound) {
@@ -2993,7 +3394,7 @@ func createHostAnonymizerForFSMode() map[string]string {
 // anonymizeHost - Add entries related to given endpoint in the host anonymizer map
 // The health report data can contain the hostname in various forms e.g. host, host:port,
 // host:port/drivepath, full url (http://host:port/drivepath)
-// The anonymizer map will have mappings for all these varients for efficiently replacing
+// The anonymizer map will have mappings for all these variants for efficiently replacing
 // any of these strings to the anonymized versions at the time of health report generation.
 func anonymizeHost(hostAnonymizer map[string]string, endpoint Endpoint, poolNum int, srvrNum int) {
 	if len(endpoint.Host) == 0 {
@@ -3041,7 +3442,7 @@ func anonymizeHost(hostAnonymizer map[string]string, endpoint Endpoint, poolNum
 	}
 }
 
-// createHostAnonymizer - Creats a map of various strings to corresponding anonymized names
+// createHostAnonymizer - Creates a map of various strings to corresponding anonymized names
 func createHostAnonymizer() map[string]string {
 	if !globalIsDistErasure {
 		return createHostAnonymizerForFSMode()

cmd/admin-handlers_test.go

@@ -168,6 +168,7 @@ func testServiceSignalReceiver(cmd cmdType, t *testing.T) {
 func getServiceCmdRequest(cmd cmdType, cred auth.Credentials) (*http.Request, error) {
 	queryVal := url.Values{}
 	queryVal.Set("action", string(cmd.toServiceAction()))
+	queryVal.Set("type", "2")
 	resource := adminPathPrefix + adminAPIVersionPrefix + "/service?" + queryVal.Encode()
 	req, err := newTestRequest(http.MethodPost, resource, 0, nil)
 	if err != nil {
@@ -220,12 +221,18 @@ func testServicesCmdHandler(cmd cmdType, t *testing.T) {
 	rec := httptest.NewRecorder()
 	adminTestBed.router.ServeHTTP(rec, req)
 
+	resp, _ := io.ReadAll(rec.Body)
 	if rec.Code != http.StatusOK {
-		resp, _ := io.ReadAll(rec.Body)
 		t.Errorf("Expected to receive %d status code but received %d. Body (%s)",
 			http.StatusOK, rec.Code, string(resp))
 	}
 
+	result := &serviceResult{}
+	if err := json.Unmarshal(resp, result); err != nil {
+		t.Error(err)
+	}
+	_ = result
+
 	// Wait until testServiceSignalReceiver() called in a goroutine quits.
 	wg.Wait()
 }

cmd/admin-heal-ops.go

@@ -28,6 +28,7 @@ import (
 	"time"
 
 	"github.com/minio/madmin-go/v3"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/logger"
 )
 
@@ -132,7 +133,13 @@ func (ahs *allHealState) popHealLocalDisks(healLocalDisks ...Endpoint) {
 func (ahs *allHealState) updateHealStatus(tracker *healingTracker) {
 	ahs.Lock()
 	defer ahs.Unlock()
-	ahs.healStatus[tracker.ID] = *tracker
+
+	tracker.mu.RLock()
+	t := *tracker
+	t.QueuedBuckets = append(make([]string, 0, len(tracker.QueuedBuckets)), tracker.QueuedBuckets...)
+	t.HealedBuckets = append(make([]string, 0, len(tracker.HealedBuckets)), tracker.HealedBuckets...)
+	ahs.healStatus[tracker.ID] = t
+	tracker.mu.RUnlock()
 }
 
 // Sort by zone, set and disk index
@@ -800,7 +807,7 @@ func (h *healSequence) healItems(objAPI ObjectLayer, bucketsOnly bool) error {
 func (h *healSequence) traverseAndHeal(objAPI ObjectLayer) {
 	bucketsOnly := false // Heals buckets and objects also.
 	h.traverseAndHealDoneCh <- h.healItems(objAPI, bucketsOnly)
-	close(h.traverseAndHealDoneCh)
+	xioutil.SafeClose(h.traverseAndHealDoneCh)
 }
 
 // healMinioSysMeta - heals all files under a given meta prefix, returns a function
@@ -810,7 +817,7 @@ func (h *healSequence) healMinioSysMeta(objAPI ObjectLayer, metaPrefix string) f
 		// NOTE: Healing on meta is run regardless
 		// of any bucket being selected, this is to ensure that
 		// meta are always upto date and correct.
-		return objAPI.HealObjects(h.ctx, minioMetaBucket, metaPrefix, h.settings, func(bucket, object, versionID string) error {
+		return objAPI.HealObjects(h.ctx, minioMetaBucket, metaPrefix, h.settings, func(bucket, object, versionID string, scanMode madmin.HealScanMode) error {
 			if h.isQuitting() {
 				return errHealStopSignalled
 			}
@@ -867,7 +874,7 @@ func (h *healSequence) healBucket(objAPI ObjectLayer, bucket string, bucketsOnly
 
 	if !h.settings.Recursive {
 		if h.object != "" {
-			if err := h.healObject(bucket, h.object, ""); err != nil {
+			if err := h.healObject(bucket, h.object, "", h.settings.ScanMode); err != nil {
 				return err
 			}
 		}
@@ -882,7 +889,7 @@ func (h *healSequence) healBucket(objAPI ObjectLayer, bucket string, bucketsOnly
 }
 
 // healObject - heal the given object and record result
-func (h *healSequence) healObject(bucket, object, versionID string) error {
+func (h *healSequence) healObject(bucket, object, versionID string, scanMode madmin.HealScanMode) error {
 	if h.isQuitting() {
 		return errHealStopSignalled
 	}

cmd/admin-router.go

@@ -154,9 +154,16 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 	}
 
 	for _, adminVersion := range adminVersions {
-		// Restart and stop MinIO service.
+		// Restart and stop MinIO service type=2
+		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/service").HandlerFunc(adminMiddleware(adminAPI.ServiceV2Handler, traceAllFlag)).Queries("action", "{action:.*}", "type", "2")
+
+		// Deprecated: Restart and stop MinIO service.
 		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/service").HandlerFunc(adminMiddleware(adminAPI.ServiceHandler, traceAllFlag)).Queries("action", "{action:.*}")
-		// Update MinIO servers.
+
+		// Update all MinIO servers type=2
+		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/update").HandlerFunc(adminMiddleware(adminAPI.ServerUpdateV2Handler, traceAllFlag)).Queries("updateURL", "{updateURL:.*}", "type", "2")
+
+		// Deprecated: Update MinIO servers.
 		adminRouter.Methods(http.MethodPost).Path(adminVersion+"/update").HandlerFunc(adminMiddleware(adminAPI.ServerUpdateHandler, traceAllFlag)).Queries("updateURL", "{updateURL:.*}")
 
 		// Info operations
@@ -300,6 +307,12 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/idp-config/{type}/{name}").HandlerFunc(adminMiddleware(adminAPI.GetIdentityProviderCfg))
 		adminRouter.Methods(http.MethodDelete).Path(adminVersion + "/idp-config/{type}/{name}").HandlerFunc(adminMiddleware(adminAPI.DeleteIdentityProviderCfg))
 
+		// LDAP specific service accounts ops
+		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/idp/ldap/add-service-account").HandlerFunc(adminMiddleware(adminAPI.AddServiceAccountLDAP))
+		adminRouter.Methods(http.MethodGet).Path(adminVersion+"/idp/ldap/list-access-keys").
+			HandlerFunc(adminMiddleware(adminAPI.ListAccessKeysLDAP)).
+			Queries("userDN", "{userDN:.*}", "listType", "{listType:.*}")
+
 		// LDAP IAM operations
 		adminRouter.Methods(http.MethodGet).Path(adminVersion + "/idp/ldap/policy-entities").HandlerFunc(adminMiddleware(adminAPI.ListLDAPPolicyMappingEntities))
 		adminRouter.Methods(http.MethodPost).Path(adminVersion + "/idp/ldap/policy/{operation}").HandlerFunc(adminMiddleware(adminAPI.AttachDetachPolicyLDAP))
@@ -376,6 +389,7 @@ func registerAdminRouter(router *mux.Router, enableConfigOps bool) {
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/peer/edit").HandlerFunc(adminMiddleware(adminAPI.SRPeerEdit))
 		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/peer/remove").HandlerFunc(adminMiddleware(adminAPI.SRPeerRemove))
 		adminRouter.Methods(http.MethodPut).Path(adminVersion+"/site-replication/resync/op").HandlerFunc(adminMiddleware(adminAPI.SiteReplicationResyncOp)).Queries("operation", "{operation:.*}")
+		adminRouter.Methods(http.MethodPut).Path(adminVersion + "/site-replication/state/edit").HandlerFunc(adminMiddleware(adminAPI.SRStateEdit))
 
 		if globalIsDistErasure {
 			// Top locks

cmd/admin-server-info.go

@@ -34,7 +34,7 @@ import (
 
 // getLocalServerProperty - returns madmin.ServerProperties for only the
 // local endpoints from given list of endpoints
-func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Request) madmin.ServerProperties {
+func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Request, metrics bool) madmin.ServerProperties {
 	addr := globalLocalNodeName
 	if r != nil {
 		addr = r.Host
@@ -141,7 +141,7 @@ func getLocalServerProperty(endpointServerPools EndpointServerPools, r *http.Req
 
 	objLayer := newObjectLayerFn()
 	if objLayer != nil {
-		storageInfo := objLayer.LocalStorageInfo(GlobalContext)
+		storageInfo := objLayer.LocalStorageInfo(GlobalContext, metrics)
 		props.State = string(madmin.ItemOnline)
 		props.Disks = storageInfo.Disks
 	} else {

cmd/api-errors.go

@@ -24,6 +24,7 @@ import (
 	"fmt"
 	"net/http"
 	"net/url"
+	"os"
 	"strconv"
 	"strings"
 
@@ -297,7 +298,6 @@ const (
 	ErrAdminConfigLDAPValidation
 	ErrAdminConfigIDPCfgNameAlreadyExists
 	ErrAdminConfigIDPCfgNameDoesNotExist
-	ErrAdminCredentialsMismatch
 	ErrInsecureClientRequest
 	ErrObjectTampered
 
@@ -310,6 +310,7 @@ const (
 	ErrSiteReplicationBucketMetaError
 	ErrSiteReplicationIAMError
 	ErrSiteReplicationConfigMissing
+	ErrSiteReplicationIAMConfigMismatch
 
 	// Pool rebalance errors
 	ErrAdminRebalanceAlreadyStarted
@@ -429,6 +430,9 @@ const (
 	ErrLambdaARNInvalid
 	ErrLambdaARNNotFound
 
+	// New Codes for GetObjectAttributes and GetObjectVersionAttributes
+	ErrInvalidAttributeName
+
 	apiErrCodeEnd // This is used only for the testing code
 )
 
@@ -955,7 +959,7 @@ var errorCodes = errorCodeMap{
 	},
 	ErrReplicationBandwidthLimitError: {
 		Code:           "XMinioAdminReplicationBandwidthLimitError",
-		Description:    "Bandwidth limit for remote target must be atleast 100MBps",
+		Description:    "Bandwidth limit for remote target must be at least 100MBps",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
 	ErrReplicationNoExistingObjects: {
@@ -1421,11 +1425,6 @@ var errorCodes = errorCodeMap{
 		Description:    "Unable to perform the requested operation because profiling is not enabled",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
-	ErrAdminCredentialsMismatch: {
-		Code:           "XMinioAdminCredentialsMismatch",
-		Description:    "Credentials in config mismatch with server environment variables",
-		HTTPStatusCode: http.StatusServiceUnavailable,
-	},
 	ErrAdminBucketQuotaExceeded: {
 		Code:           "XMinioAdminBucketQuotaExceeded",
 		Description:    "Bucket quota exceeded",
@@ -1512,6 +1511,11 @@ var errorCodes = errorCodeMap{
 		Description:    "Site not found in site replication configuration",
 		HTTPStatusCode: http.StatusBadRequest,
 	},
+	ErrSiteReplicationIAMConfigMismatch: {
+		Code:           "XMinioSiteReplicationIAMConfigMismatch",
+		Description:    "IAM configuration mismatch between sites",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	ErrAdminRebalanceAlreadyStarted: {
 		Code:           "XMinioAdminRebalanceAlreadyStarted",
 		Description:    "Pool rebalance is already started",
@@ -2057,6 +2061,11 @@ var errorCodes = errorCodeMap{
 		Description:    "The specified policy is not found.",
 		HTTPStatusCode: http.StatusNotFound,
 	},
+	ErrInvalidAttributeName: {
+		Code:           "InvalidArgument",
+		Description:    "Invalid attribute name specified.",
+		HTTPStatusCode: http.StatusBadRequest,
+	},
 	// Add your error structure here.
 }
 
@@ -2068,6 +2077,11 @@ func toAPIErrorCode(ctx context.Context, err error) (apiErr APIErrorCode) {
 		return ErrNone
 	}
 
+	// Errors that are generated by net.Conn and any context errors must be handled here.
+	if errors.Is(err, os.ErrDeadlineExceeded) || errors.Is(err, context.DeadlineExceeded) {
+		return ErrRequestTimedout
+	}
+
 	// Only return ErrClientDisconnected if the provided context is actually canceled.
 	// This way downstream context.Canceled will still report ErrRequestTimedout
 	if contextCanceled(ctx) && errors.Is(ctx.Err(), context.Canceled) {
@@ -2337,25 +2351,10 @@ func toAPIErrorCode(ctx context.Context, err error) (apiErr APIErrorCode) {
 	case dns.ErrBucketConflict:
 		apiErr = ErrBucketAlreadyExists
 	default:
-		var ie, iw int
-		// This work-around is to handle the issue golang/go#30648
-		//nolint:gocritic
-		if _, ferr := fmt.Fscanf(strings.NewReader(err.Error()),
-			"request declared a Content-Length of %d but only wrote %d bytes",
-			&ie, &iw); ferr != nil {
-			apiErr = ErrInternalError
-			// Make sure to log the errors which we cannot translate
-			// to a meaningful S3 API errors. This is added to aid in
-			// debugging unexpected/unhandled errors.
-			logger.LogIf(ctx, err)
-		} else if ie > iw {
+		if strings.Contains(err.Error(), "request declared a Content-Length") {
 			apiErr = ErrIncompleteBody
 		} else {
 			apiErr = ErrInternalError
-			// Make sure to log the errors which we cannot translate
-			// to a meaningful S3 API errors. This is added to aid in
-			// debugging unexpected/unhandled errors.
-			logger.LogIf(ctx, err)
 		}
 	}
 
@@ -2499,6 +2498,13 @@ func toAPIError(ctx context.Context, err error) APIError {
 		}
 	}
 
+	if apiErr.Code == "InternalError" {
+		// Make sure to log the errors which we cannot translate
+		// to a meaningful S3 API errors. This is added to aid in
+		// debugging unexpected/unhandled errors.
+		logger.LogIf(ctx, err)
+	}
+
 	return apiErr
 }
 

cmd/api-headers.go

@@ -133,11 +133,6 @@ func setObjectHeaders(w http.ResponseWriter, objInfo ObjectInfo, rs *HTTPRangeSp
 		w.Header().Set(xhttp.Expires, objInfo.Expires.UTC().Format(http.TimeFormat))
 	}
 
-	if globalCacheConfig.Enabled {
-		w.Header().Set(xhttp.XCache, objInfo.CacheStatus.String())
-		w.Header().Set(xhttp.XCacheLookup, objInfo.CacheLookupStatus.String())
-	}
-
 	// Set tag count if object has tags
 	if len(objInfo.UserTags) > 0 {
 		tags, _ := tags.ParseObjectTags(objInfo.UserTags)

cmd/api-response.go

@@ -502,6 +502,47 @@ func generateListBucketsResponse(buckets []BucketInfo) ListBucketsResponse {
 	return data
 }
 
+func cleanReservedKeys(metadata map[string]string) map[string]string {
+	m := cloneMSS(metadata)
+
+	switch kind, _ := crypto.IsEncrypted(metadata); kind {
+	case crypto.S3:
+		m[xhttp.AmzServerSideEncryption] = xhttp.AmzEncryptionAES
+	case crypto.S3KMS:
+		m[xhttp.AmzServerSideEncryption] = xhttp.AmzEncryptionKMS
+		m[xhttp.AmzServerSideEncryptionKmsID] = kmsKeyIDFromMetadata(metadata)
+		if kmsCtx, ok := metadata[crypto.MetaContext]; ok {
+			m[xhttp.AmzServerSideEncryptionKmsContext] = kmsCtx
+		}
+	case crypto.SSEC:
+		m[xhttp.AmzServerSideEncryptionCustomerAlgorithm] = xhttp.AmzEncryptionAES
+
+	}
+
+	var toRemove []string
+	for k := range cleanMinioInternalMetadataKeys(m) {
+		if stringsHasPrefixFold(k, ReservedMetadataPrefixLower) {
+			// Do not need to send any internal metadata
+			// values to client.
+			toRemove = append(toRemove, k)
+			continue
+		}
+
+		// https://github.com/google/security-research/security/advisories/GHSA-76wf-9vgp-pj7w
+		if equals(k, xhttp.AmzMetaUnencryptedContentLength, xhttp.AmzMetaUnencryptedContentMD5) {
+			toRemove = append(toRemove, k)
+			continue
+		}
+	}
+
+	for _, k := range toRemove {
+		delete(m, k)
+		delete(m, strings.ToLower(k))
+	}
+
+	return m
+}
+
 // generates an ListBucketVersions response for the said bucket with other enumerated options.
 func generateListVersionsResponse(bucket, prefix, marker, versionIDMarker, delimiter, encodingType string, maxKeys int, resp ListObjectVersionsInfo, metadata metaCheckFn) ListVersionsResponse {
 	versions := make([]ObjectVersion, 0, len(resp.Objects))
@@ -549,18 +590,11 @@ func generateListVersionsResponse(bucket, prefix, marker, versionIDMarker, delim
 			case crypto.SSEC:
 				content.UserMetadata.Set(xhttp.AmzServerSideEncryptionCustomerAlgorithm, xhttp.AmzEncryptionAES)
 			}
-			for k, v := range cleanMinioInternalMetadataKeys(object.UserDefined) {
-				if stringsHasPrefixFold(k, ReservedMetadataPrefixLower) {
-					// Do not need to send any internal metadata
-					// values to client.
-					continue
-				}
-				// https://github.com/google/security-research/security/advisories/GHSA-76wf-9vgp-pj7w
-				if equals(k, xhttp.AmzMetaUnencryptedContentLength, xhttp.AmzMetaUnencryptedContentMD5) {
-					continue
-				}
+			for k, v := range cleanReservedKeys(object.UserDefined) {
 				content.UserMetadata.Set(k, v)
 			}
+
+			content.UserMetadata.Set("expires", object.Expires.Format(http.TimeFormat))
 			content.Internal = &ObjectInternalInfo{
 				K: object.DataBlocks,
 				M: object.ParityBlocks,
@@ -692,18 +726,10 @@ func generateListObjectsV2Response(bucket, prefix, token, nextToken, startAfter,
 				case crypto.SSEC:
 					content.UserMetadata.Set(xhttp.AmzServerSideEncryptionCustomerAlgorithm, xhttp.AmzEncryptionAES)
 				}
-				for k, v := range cleanMinioInternalMetadataKeys(object.UserDefined) {
-					if stringsHasPrefixFold(k, ReservedMetadataPrefixLower) {
-						// Do not need to send any internal metadata
-						// values to client.
-						continue
-					}
-					// https://github.com/google/security-research/security/advisories/GHSA-76wf-9vgp-pj7w
-					if equals(k, xhttp.AmzMetaUnencryptedContentLength, xhttp.AmzMetaUnencryptedContentMD5) {
-						continue
-					}
+				for k, v := range cleanReservedKeys(object.UserDefined) {
 					content.UserMetadata.Set(k, v)
 				}
+				content.UserMetadata.Set("expires", object.Expires.Format(http.TimeFormat))
 				content.Internal = &ObjectInternalInfo{
 					K: object.DataBlocks,
 					M: object.ParityBlocks,
@@ -920,11 +946,13 @@ func writeSuccessResponseHeadersOnly(w http.ResponseWriter) {
 
 // writeErrorRespone writes error headers
 func writeErrorResponse(ctx context.Context, w http.ResponseWriter, err APIError, reqURL *url.URL) {
-	switch err.Code {
-	case "SlowDown", "XMinioServerNotInitialized", "XMinioReadQuorum", "XMinioWriteQuorum":
+	if err.HTTPStatusCode == http.StatusServiceUnavailable {
 		// Set retry-after header to indicate user-agents to retry request after 120secs.
 		// https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Retry-After
 		w.Header().Set(xhttp.RetryAfter, "120")
+	}
+
+	switch err.Code {
 	case "InvalidRegion":
 		err.Description = fmt.Sprintf("Region does not match; expecting '%s'.", globalSite.Region)
 	case "AuthorizationHeaderMalformed":
@@ -978,7 +1006,7 @@ func writeCustomErrorResponseJSON(ctx context.Context, w http.ResponseWriter, er
 		BucketName: reqInfo.BucketName,
 		Key:        reqInfo.ObjectName,
 		RequestID:  w.Header().Get(xhttp.AmzRequestID),
-		HostID:     globalDeploymentID,
+		HostID:     globalDeploymentID(),
 	}
 	encodedErrorResponse := encodeResponseJSON(errorResponse)
 	writeResponse(w, err.HTTPStatusCode, encodedErrorResponse, mimeJSON)

cmd/api-router.go

@@ -23,7 +23,7 @@ import (
 	"net/http"
 
 	"github.com/klauspost/compress/gzhttp"
-	"github.com/minio/console/restapi"
+	consoleapi "github.com/minio/console/api"
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/mux"
@@ -43,13 +43,13 @@ func setHTTPServer(h *xhttp.Server) {
 	globalObjLayerMutex.Unlock()
 }
 
-func newConsoleServerFn() *restapi.Server {
+func newConsoleServerFn() *consoleapi.Server {
 	globalObjLayerMutex.RLock()
 	defer globalObjLayerMutex.RUnlock()
 	return globalConsoleSrv
 }
 
-func setConsoleSrv(srv *restapi.Server) {
+func setConsoleSrv(srv *consoleapi.Server) {
 	globalObjLayerMutex.Lock()
 	globalConsoleSrv = srv
 	globalObjLayerMutex.Unlock()
@@ -61,18 +61,6 @@ func newObjectLayerFn() ObjectLayer {
 	return globalObjectAPI
 }
 
-func newCachedObjectLayerFn() CacheObjectLayer {
-	globalObjLayerMutex.RLock()
-	defer globalObjLayerMutex.RUnlock()
-	return globalCacheObjectAPI
-}
-
-func setCacheObjectLayer(c CacheObjectLayer) {
-	globalObjLayerMutex.Lock()
-	globalCacheObjectAPI = c
-	globalObjLayerMutex.Unlock()
-}
-
 func setObjectLayer(o ObjectLayer) {
 	globalObjLayerMutex.Lock()
 	globalObjectAPI = o
@@ -82,7 +70,6 @@ func setObjectLayer(o ObjectLayer) {
 // objectAPIHandler implements and provides http handlers for S3 API.
 type objectAPIHandlers struct {
 	ObjectAPI func() ObjectLayer
-	CacheAPI  func() CacheObjectLayer
 }
 
 // getHost tries its best to return the request host.
@@ -189,7 +176,6 @@ func registerAPIRouter(router *mux.Router) {
 	// Initialize API.
 	api := objectAPIHandlers{
 		ObjectAPI: newObjectLayerFn,
-		CacheAPI:  newCachedObjectLayerFn,
 	}
 
 	// API Router
@@ -241,6 +227,9 @@ func registerAPIRouter(router *mux.Router) {
 		// HeadObject
 		router.Methods(http.MethodHead).Path("/{object:.+}").HandlerFunc(
 			collectAPIStats("headobject", maxClients(gz(httpTraceAll(api.HeadObjectHandler)))))
+		// GetObjectAttributes
+		router.Methods(http.MethodGet).Path("/{object:.+}").HandlerFunc(
+			collectAPIStats("getobjectattributes", maxClients(gz(httpTraceHdrs(api.GetObjectAttributesHandler))))).Queries("attributes", "")
 		// CopyObjectPart
 		router.Methods(http.MethodPut).Path("/{object:.+}").
 			HeadersRegexp(xhttp.AmzCopySource, ".*?(\\/|%2F).*?").

cmd/auth-handler.go

@@ -162,7 +162,8 @@ func validateAdminSignature(ctx context.Context, r *http.Request, region string)
 	s3Err := ErrAccessDenied
 	if _, ok := r.Header[xhttp.AmzContentSha256]; ok &&
 		getRequestAuthType(r) == authTypeSigned {
-		// We only support admin credentials to access admin APIs.
+
+		// Get credential information from the request.
 		cred, owner, s3Err = getReqAccessKeyV4(r, region, serviceS3)
 		if s3Err != ErrNone {
 			return cred, owner, s3Err
@@ -256,7 +257,7 @@ func getClaimsFromTokenWithSecret(token, secret string) (map[string]interface{},
 		if err != nil {
 			// Base64 decoding fails, we should log to indicate
 			// something is malforming the request sent by client.
-			logger.LogIf(GlobalContext, err, logger.Application)
+			logger.LogIf(GlobalContext, err, logger.ErrorKind)
 			return nil, errAuthentication
 		}
 		claims.MapClaims[sessionPolicyNameExtracted] = string(spBytes)
@@ -338,7 +339,7 @@ func checkRequestAuthTypeWithVID(ctx context.Context, r *http.Request, action po
 
 func authenticateRequest(ctx context.Context, r *http.Request, action policy.Action) (s3Err APIErrorCode) {
 	if logger.GetReqInfo(ctx) == nil {
-		logger.LogIf(ctx, errors.New("unexpected context.Context does not have a logger.ReqInfo"), logger.Minio)
+		logger.LogIf(ctx, errors.New("unexpected context.Context does not have a logger.ReqInfo"), logger.ErrorKind)
 		return ErrAccessDenied
 	}
 
@@ -377,7 +378,7 @@ func authenticateRequest(ctx context.Context, r *http.Request, action policy.Act
 		// To extract region from XML in request body, get copy of request body.
 		payload, err := io.ReadAll(io.LimitReader(r.Body, maxLocationConstraintSize))
 		if err != nil {
-			logger.LogIf(ctx, err, logger.Application)
+			logger.LogIf(ctx, err, logger.ErrorKind)
 			return ErrMalformedXML
 		}
 

cmd/auth-handler_test.go

@@ -287,7 +287,7 @@ func mustNewSignedRequest(method string, urlStr string, contentLength int64, bod
 	req := mustNewRequest(method, urlStr, contentLength, body, t)
 	cred := globalActiveCred
 	if err := signRequestV4(req, cred.AccessKey, cred.SecretKey); err != nil {
-		t.Fatalf("Unable to inititalized new signed http request %s", err)
+		t.Fatalf("Unable to initialized new signed http request %s", err)
 	}
 	return req
 }
@@ -298,7 +298,7 @@ func mustNewSignedV2Request(method string, urlStr string, contentLength int64, b
 	req := mustNewRequest(method, urlStr, contentLength, body, t)
 	cred := globalActiveCred
 	if err := signRequestV2(req, cred.AccessKey, cred.SecretKey); err != nil {
-		t.Fatalf("Unable to inititalized new signed http request %s", err)
+		t.Fatalf("Unable to initialized new signed http request %s", err)
 	}
 	return req
 }
@@ -309,7 +309,7 @@ func mustNewPresignedV2Request(method string, urlStr string, contentLength int64
 	req := mustNewRequest(method, urlStr, contentLength, body, t)
 	cred := globalActiveCred
 	if err := preSignV2(req, cred.AccessKey, cred.SecretKey, time.Now().Add(10*time.Minute).Unix()); err != nil {
-		t.Fatalf("Unable to inititalized new signed http request %s", err)
+		t.Fatalf("Unable to initialized new signed http request %s", err)
 	}
 	return req
 }
@@ -320,7 +320,7 @@ func mustNewPresignedRequest(method string, urlStr string, contentLength int64,
 	req := mustNewRequest(method, urlStr, contentLength, body, t)
 	cred := globalActiveCred
 	if err := preSignV4(req, cred.AccessKey, cred.SecretKey, time.Now().Add(10*time.Minute).Unix()); err != nil {
-		t.Fatalf("Unable to inititalized new signed http request %s", err)
+		t.Fatalf("Unable to initialized new signed http request %s", err)
 	}
 	return req
 }
@@ -491,7 +491,7 @@ func TestValidateAdminSignature(t *testing.T) {
 	for i, testCase := range testCases {
 		req := mustNewRequest(http.MethodGet, "http://localhost:9000/", 0, nil, t)
 		if err := signRequestV4(req, testCase.AccessKey, testCase.SecretKey); err != nil {
-			t.Fatalf("Unable to inititalized new signed http request %s", err)
+			t.Fatalf("Unable to initialized new signed http request %s", err)
 		}
 		_, _, s3Error := validateAdminSignature(ctx, req, globalMinioDefaultRegion)
 		if s3Error != testCase.ErrCode {

cmd/background-newdisks-heal-ops.go

@@ -232,7 +232,7 @@ func (h *healingTracker) delete(ctx context.Context) error {
 		pathJoin(bucketMetaPrefix, healingTrackerFilename),
 		DeleteOptions{
 			Recursive: false,
-			Force:     false,
+			Immediate: false,
 		},
 	)
 }
@@ -345,18 +345,17 @@ func initAutoHeal(ctx context.Context, objAPI ObjectLayer) {
 
 	initBackgroundHealing(ctx, objAPI) // start quick background healing
 
-	globalBackgroundHealState.pushHealLocalDisks(getLocalDisksToHeal()...)
-
 	if env.Get("_MINIO_AUTO_DRIVE_HEALING", config.EnableOn) == config.EnableOn || env.Get("_MINIO_AUTO_DISK_HEALING", config.EnableOn) == config.EnableOn {
+		globalBackgroundHealState.pushHealLocalDisks(getLocalDisksToHeal()...)
 		go monitorLocalDisksAndHeal(ctx, z)
 	}
 }
 
 func getLocalDisksToHeal() (disksToHeal Endpoints) {
 	globalLocalDrivesMu.RLock()
-	globalLocalDrives := globalLocalDrives
+	localDrives := globalLocalDrives
 	globalLocalDrivesMu.RUnlock()
-	for _, disk := range globalLocalDrives {
+	for _, disk := range localDrives {
 		_, err := disk.GetDiskID()
 		if errors.Is(err, errUnformattedDisk) {
 			disksToHeal = append(disksToHeal, disk.Endpoint())
@@ -377,7 +376,7 @@ func getLocalDisksToHeal() (disksToHeal Endpoints) {
 var newDiskHealingTimeout = newDynamicTimeout(30*time.Second, 10*time.Second)
 
 func healFreshDisk(ctx context.Context, z *erasureServerPools, endpoint Endpoint) error {
-	disk, format, err := connectEndpoint(endpoint)
+	disk, format, _, err := connectEndpoint(endpoint)
 	if err != nil {
 		return fmt.Errorf("Error: %w, %s", err, endpoint)
 	}
@@ -491,6 +490,10 @@ func healFreshDisk(ctx context.Context, z *erasureServerPools, endpoint Endpoint
 	}
 
 	for _, disk := range disks {
+		if disk == nil {
+			continue
+		}
+
 		t, err := loadHealingTracker(ctx, disk)
 		if err != nil {
 			if !errors.Is(err, errFileNotFound) {

cmd/batch-expire.go

@@ -0,0 +1,746 @@
+// Copyright (c) 2015-2023 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 (
+	"bytes"
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"runtime"
+	"strconv"
+	"time"
+
+	"github.com/minio/minio-go/v7/pkg/tags"
+	"github.com/minio/minio/internal/bucket/versioning"
+	xhttp "github.com/minio/minio/internal/http"
+	xioutil "github.com/minio/minio/internal/ioutil"
+	"github.com/minio/minio/internal/logger"
+	"github.com/minio/pkg/v2/env"
+	"github.com/minio/pkg/v2/wildcard"
+	"github.com/minio/pkg/v2/workers"
+	"gopkg.in/yaml.v3"
+)
+
+// expire: # Expire objects that match a condition
+//   apiVersion: v1
+//   bucket: mybucket # Bucket where this batch job will expire matching objects from
+//   prefix: myprefix # (Optional) Prefix under which this job will expire objects matching the rules below.
+//   rules:
+//     - type: object  # regular objects with zero or more older versions
+//       name: NAME # match object names that satisfy the wildcard expression.
+//       olderThan: 70h # match objects older than this value
+//       createdBefore: "2006-01-02T15:04:05.00Z" # match objects created before "date"
+//       tags:
+//         - key: name
+//           value: pick* # match objects with tag 'name', all values starting with 'pick'
+//       metadata:
+//         - key: content-type
+//           value: image/* # match objects with 'content-type', all values starting with 'image/'
+//       size:
+//         lessThan: "10MiB" # match objects with size less than this value (e.g. 10MiB)
+//         greaterThan: 1MiB # match objects with size greater than this value (e.g. 1MiB)
+//       purge:
+//           # retainVersions: 0 # (default) delete all versions of the object. This option is the fastest.
+//           # retainVersions: 5 # keep the latest 5 versions of the object.
+//
+//     - type: deleted # objects with delete marker as their latest version
+//       name: NAME # match object names that satisfy the wildcard expression.
+//       olderThan: 10h # match objects older than this value (e.g. 7d10h31s)
+//       createdBefore: "2006-01-02T15:04:05.00Z" # match objects created before "date"
+//       purge:
+//           # retainVersions: 0 # (default) delete all versions of the object. This option is the fastest.
+//           # retainVersions: 5 # keep the latest 5 versions of the object including delete markers.
+//
+//   notify:
+//     endpoint: https://notify.endpoint # notification endpoint to receive job completion status
+//     token: Bearer xxxxx # optional authentication token for the notification endpoint
+//
+//   retry:
+//     attempts: 10 # number of retries for the job before giving up
+//     delay: 500ms # least amount of delay between each retry
+
+//go:generate msgp -file $GOFILE
+
+// BatchJobExpirePurge type accepts non-negative versions to be retained
+type BatchJobExpirePurge struct {
+	line, col      int
+	RetainVersions int `yaml:"retainVersions" json:"retainVersions"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobExpirePurge{}
+
+// UnmarshalYAML - BatchJobExpirePurge extends unmarshal to extract line, col
+func (p *BatchJobExpirePurge) UnmarshalYAML(val *yaml.Node) error {
+	type purge BatchJobExpirePurge
+	var tmp purge
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*p = BatchJobExpirePurge(tmp)
+	p.line, p.col = val.Line, val.Column
+	return nil
+}
+
+// Validate returns nil if value is valid, ie > 0.
+func (p BatchJobExpirePurge) Validate() error {
+	if p.RetainVersions < 0 {
+		return BatchJobYamlErr{
+			line: p.line,
+			col:  p.col,
+			msg:  "retainVersions must be >= 0",
+		}
+	}
+	return nil
+}
+
+// BatchJobExpireFilter holds all the filters currently supported for batch replication
+type BatchJobExpireFilter struct {
+	line, col     int
+	OlderThan     time.Duration       `yaml:"olderThan,omitempty" json:"olderThan"`
+	CreatedBefore *time.Time          `yaml:"createdBefore,omitempty" json:"createdBefore"`
+	Tags          []BatchJobKV        `yaml:"tags,omitempty" json:"tags"`
+	Metadata      []BatchJobKV        `yaml:"metadata,omitempty" json:"metadata"`
+	Size          BatchJobSizeFilter  `yaml:"size" json:"size"`
+	Type          string              `yaml:"type" json:"type"`
+	Name          string              `yaml:"name" json:"name"`
+	Purge         BatchJobExpirePurge `yaml:"purge" json:"purge"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobExpireFilter{}
+
+// UnmarshalYAML - BatchJobExpireFilter extends unmarshal to extract line, col
+// information
+func (ef *BatchJobExpireFilter) UnmarshalYAML(value *yaml.Node) error {
+	type expFilter BatchJobExpireFilter
+	var tmp expFilter
+	err := value.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+	*ef = BatchJobExpireFilter(tmp)
+	ef.line, ef.col = value.Line, value.Column
+	return err
+}
+
+// Matches returns true if obj matches the filter conditions specified in ef.
+func (ef BatchJobExpireFilter) Matches(obj ObjectInfo, now time.Time) bool {
+	switch ef.Type {
+	case BatchJobExpireObject:
+		if obj.DeleteMarker {
+			return false
+		}
+	case BatchJobExpireDeleted:
+		if !obj.DeleteMarker {
+			return false
+		}
+	default:
+		// we should never come here, Validate should have caught this.
+		logger.LogOnceIf(context.Background(), fmt.Errorf("invalid filter type: %s", ef.Type), ef.Type)
+		return false
+	}
+
+	if len(ef.Name) > 0 && !wildcard.Match(ef.Name, obj.Name) {
+		return false
+	}
+	if ef.OlderThan > 0 && now.Sub(obj.ModTime) <= ef.OlderThan {
+		return false
+	}
+
+	if ef.CreatedBefore != nil && !obj.ModTime.Before(*ef.CreatedBefore) {
+		return false
+	}
+
+	if len(ef.Tags) > 0 && !obj.DeleteMarker {
+		// Only parse object tags if tags filter is specified.
+		var tagMap map[string]string
+		if len(obj.UserTags) != 0 {
+			t, err := tags.ParseObjectTags(obj.UserTags)
+			if err != nil {
+				return false
+			}
+			tagMap = t.ToMap()
+		}
+
+		for _, kv := range ef.Tags {
+			// Object (version) must match all tags specified in
+			// the filter
+			var match bool
+			for t, v := range tagMap {
+				if kv.Match(BatchJobKV{Key: t, Value: v}) {
+					match = true
+				}
+			}
+			if !match {
+				return false
+			}
+		}
+
+	}
+	if len(ef.Metadata) > 0 && !obj.DeleteMarker {
+		for _, kv := range ef.Metadata {
+			// Object (version) must match all x-amz-meta and
+			// standard metadata headers
+			// specified in the filter
+			var match bool
+			for k, v := range obj.UserDefined {
+				if !stringsHasPrefixFold(k, "x-amz-meta-") && !isStandardHeader(k) {
+					continue
+				}
+				// We only need to match x-amz-meta or standardHeaders
+				if kv.Match(BatchJobKV{Key: k, Value: v}) {
+					match = true
+				}
+			}
+			if !match {
+				return false
+			}
+		}
+	}
+
+	return ef.Size.InRange(obj.Size)
+}
+
+const (
+	// BatchJobExpireObject - object type
+	BatchJobExpireObject string = "object"
+	// BatchJobExpireDeleted - delete marker type
+	BatchJobExpireDeleted string = "deleted"
+)
+
+// Validate returns nil if ef has valid fields, validation error otherwise.
+func (ef BatchJobExpireFilter) Validate() error {
+	switch ef.Type {
+	case BatchJobExpireObject:
+	case BatchJobExpireDeleted:
+		if len(ef.Tags) > 0 || len(ef.Metadata) > 0 {
+			return BatchJobYamlErr{
+				line: ef.line,
+				col:  ef.col,
+				msg:  "delete type filter can't have tags or metadata",
+			}
+		}
+	default:
+		return BatchJobYamlErr{
+			line: ef.line,
+			col:  ef.col,
+			msg:  "invalid batch-expire type",
+		}
+	}
+
+	for _, tag := range ef.Tags {
+		if err := tag.Validate(); err != nil {
+			return err
+		}
+	}
+
+	for _, meta := range ef.Metadata {
+		if err := meta.Validate(); err != nil {
+			return err
+		}
+	}
+	if err := ef.Purge.Validate(); err != nil {
+		return err
+	}
+	if err := ef.Size.Validate(); err != nil {
+		return err
+	}
+	if ef.CreatedBefore != nil && !ef.CreatedBefore.Before(time.Now()) {
+		return BatchJobYamlErr{
+			line: ef.line,
+			col:  ef.col,
+			msg:  "CreatedBefore is in the future",
+		}
+	}
+	return nil
+}
+
+// BatchJobExpire represents configuration parameters for a batch expiration
+// job typically supplied in yaml form
+type BatchJobExpire struct {
+	line, col       int
+	APIVersion      string                 `yaml:"apiVersion" json:"apiVersion"`
+	Bucket          string                 `yaml:"bucket" json:"bucket"`
+	Prefix          string                 `yaml:"prefix" json:"prefix"`
+	NotificationCfg BatchJobNotification   `yaml:"notify" json:"notify"`
+	Retry           BatchJobRetry          `yaml:"retry" json:"retry"`
+	Rules           []BatchJobExpireFilter `yaml:"rules" json:"rules"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobExpire{}
+
+// UnmarshalYAML - BatchJobExpire extends default unmarshal to extract line, col information.
+func (r *BatchJobExpire) UnmarshalYAML(val *yaml.Node) error {
+	type expireJob BatchJobExpire
+	var tmp expireJob
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*r = BatchJobExpire(tmp)
+	r.line, r.col = val.Line, val.Column
+	return nil
+}
+
+// Notify notifies notification endpoint if configured regarding job failure or success.
+func (r BatchJobExpire) Notify(ctx context.Context, body io.Reader) error {
+	if r.NotificationCfg.Endpoint == "" {
+		return nil
+	}
+
+	ctx, cancel := context.WithTimeout(ctx, 10*time.Second)
+	defer cancel()
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, r.NotificationCfg.Endpoint, body)
+	if err != nil {
+		return err
+	}
+
+	if r.NotificationCfg.Token != "" {
+		req.Header.Set("Authorization", r.NotificationCfg.Token)
+	}
+
+	clnt := http.Client{Transport: getRemoteInstanceTransport}
+	resp, err := clnt.Do(req)
+	if err != nil {
+		return err
+	}
+
+	xhttp.DrainBody(resp.Body)
+	if resp.StatusCode != http.StatusOK {
+		return errors.New(resp.Status)
+	}
+
+	return nil
+}
+
+// Expire expires object versions which have already matched supplied filter conditions
+func (r *BatchJobExpire) Expire(ctx context.Context, api ObjectLayer, vc *versioning.Versioning, objsToDel []ObjectToDelete) []error {
+	opts := ObjectOptions{
+		PrefixEnabledFn:  vc.PrefixEnabled,
+		VersionSuspended: vc.Suspended(),
+	}
+	_, errs := api.DeleteObjects(ctx, r.Bucket, objsToDel, opts)
+	return errs
+}
+
+const (
+	batchExpireName                 = "batch-expire.bin"
+	batchExpireFormat               = 1
+	batchExpireVersionV1            = 1
+	batchExpireVersion              = batchExpireVersionV1
+	batchExpireAPIVersion           = "v1"
+	batchExpireJobDefaultRetries    = 3
+	batchExpireJobDefaultRetryDelay = 250 * time.Millisecond
+)
+
+type objInfoCache map[string]*ObjectInfo
+
+func newObjInfoCache() objInfoCache {
+	return objInfoCache(make(map[string]*ObjectInfo))
+}
+
+func (oiCache objInfoCache) Add(toDel ObjectToDelete, oi *ObjectInfo) {
+	oiCache[fmt.Sprintf("%s-%s", toDel.ObjectName, toDel.VersionID)] = oi
+}
+
+func (oiCache objInfoCache) Get(toDel ObjectToDelete) (*ObjectInfo, bool) {
+	oi, ok := oiCache[fmt.Sprintf("%s-%s", toDel.ObjectName, toDel.VersionID)]
+	return oi, ok
+}
+
+func batchObjsForDelete(ctx context.Context, r *BatchJobExpire, ri *batchJobInfo, job BatchJobRequest, api ObjectLayer, wk *workers.Workers, expireCh <-chan []expireObjInfo) {
+	vc, _ := globalBucketVersioningSys.Get(r.Bucket)
+	retryAttempts := r.Retry.Attempts
+	delay := job.Expire.Retry.Delay
+	if delay == 0 {
+		delay = batchExpireJobDefaultRetryDelay
+	}
+
+	var i int
+	for toExpire := range expireCh {
+		select {
+		case <-ctx.Done():
+			return
+		default:
+		}
+		if i > 0 {
+			if wait := globalBatchConfig.ExpirationWait(); wait > 0 {
+				time.Sleep(wait)
+			}
+		}
+		i++
+		wk.Take()
+		go func(toExpire []expireObjInfo) {
+			defer wk.Give()
+
+			toExpireAll := make([]ObjectInfo, 0, len(toExpire))
+			toDel := make([]ObjectToDelete, 0, len(toExpire))
+			oiCache := newObjInfoCache()
+			for _, exp := range toExpire {
+				if exp.ExpireAll {
+					toExpireAll = append(toExpireAll, exp.ObjectInfo)
+					continue
+				}
+				// Cache ObjectInfo value via pointers for
+				// subsequent use to track objects which
+				// couldn't be deleted.
+				od := ObjectToDelete{
+					ObjectV: ObjectV{
+						ObjectName: exp.Name,
+						VersionID:  exp.VersionID,
+					},
+				}
+				toDel = append(toDel, od)
+				oiCache.Add(od, &exp.ObjectInfo)
+			}
+
+			var done bool
+			// DeleteObject(deletePrefix: true) to expire all versions of an object
+			for _, exp := range toExpireAll {
+				var success bool
+				for attempts := 1; attempts <= retryAttempts; attempts++ {
+					select {
+					case <-ctx.Done():
+						done = true
+					default:
+					}
+					stopFn := globalBatchJobsMetrics.trace(batchJobMetricExpire, ri.JobID, attempts)
+					_, err := api.DeleteObject(ctx, exp.Bucket, exp.Name, ObjectOptions{
+						DeletePrefix: true,
+					})
+					if err != nil {
+						stopFn(exp, err)
+						logger.LogIf(ctx, fmt.Errorf("Failed to expire %s/%s versionID=%s due to %v (attempts=%d)", toExpire[i].Bucket, toExpire[i].Name, toExpire[i].VersionID, err, attempts))
+					} else {
+						stopFn(exp, err)
+						success = true
+						break
+					}
+				}
+				ri.trackMultipleObjectVersions(r.Bucket, exp, success)
+				if done {
+					break
+				}
+			}
+
+			if done {
+				return
+			}
+
+			// DeleteMultiple objects
+			toDelCopy := make([]ObjectToDelete, len(toDel))
+			for attempts := 1; attempts <= retryAttempts; attempts++ {
+				select {
+				case <-ctx.Done():
+					return
+				default:
+				}
+
+				stopFn := globalBatchJobsMetrics.trace(batchJobMetricExpire, ri.JobID, attempts)
+				// Copying toDel to select from objects whose
+				// deletion failed
+				copy(toDelCopy, toDel)
+				var failed int
+				errs := r.Expire(ctx, api, vc, toDel)
+				// reslice toDel in preparation for next retry
+				// attempt
+				toDel = toDel[:0]
+				for i, err := range errs {
+					if err != nil {
+						stopFn(toDelCopy[i], err)
+						logger.LogIf(ctx, fmt.Errorf("Failed to expire %s/%s versionID=%s due to %v (attempts=%d)", ri.Bucket, toDelCopy[i].ObjectName, toDelCopy[i].VersionID, err, attempts))
+						failed++
+						if attempts == retryAttempts { // all retry attempts failed, record failure
+							if oi, ok := oiCache.Get(toDelCopy[i]); ok {
+								ri.trackCurrentBucketObject(r.Bucket, *oi, false)
+							}
+						} else {
+							toDel = append(toDel, toDelCopy[i])
+						}
+					} else {
+						stopFn(toDelCopy[i], nil)
+						if oi, ok := oiCache.Get(toDelCopy[i]); ok {
+							ri.trackCurrentBucketObject(r.Bucket, *oi, true)
+						}
+					}
+				}
+
+				globalBatchJobsMetrics.save(ri.JobID, ri)
+
+				if failed == 0 {
+					break
+				}
+
+				// Add a delay between retry attempts
+				if attempts < retryAttempts {
+					time.Sleep(delay)
+				}
+			}
+		}(toExpire)
+	}
+}
+
+type expireObjInfo struct {
+	ObjectInfo
+	ExpireAll bool
+}
+
+// Start the batch expiration job, resumes if there was a pending job via "job.ID"
+func (r *BatchJobExpire) Start(ctx context.Context, api ObjectLayer, job BatchJobRequest) error {
+	ri := &batchJobInfo{
+		JobID:     job.ID,
+		JobType:   string(job.Type()),
+		StartTime: job.Started,
+	}
+	if err := ri.load(ctx, api, job); err != nil {
+		return err
+	}
+
+	globalBatchJobsMetrics.save(job.ID, ri)
+	lastObject := ri.Object
+
+	now := time.Now().UTC()
+
+	workerSize, err := strconv.Atoi(env.Get("_MINIO_BATCH_EXPIRATION_WORKERS", strconv.Itoa(runtime.GOMAXPROCS(0)/2)))
+	if err != nil {
+		return err
+	}
+
+	wk, err := workers.New(workerSize)
+	if err != nil {
+		// invalid worker size.
+		return err
+	}
+
+	ctx, cancel := context.WithCancel(ctx)
+	defer cancel()
+
+	results := make(chan ObjectInfo, workerSize)
+	if err := api.Walk(ctx, r.Bucket, r.Prefix, results, WalkOptions{
+		Marker:       lastObject,
+		LatestOnly:   false, // we need to visit all versions of the object to implement purge: retainVersions
+		VersionsSort: WalkVersionsSortDesc,
+	}); err != nil {
+		// Do not need to retry if we can't list objects on source.
+		return err
+	}
+
+	// Goroutine to periodically save batch-expire job's in-memory state
+	saverQuitCh := make(chan struct{})
+	go func() {
+		saveTicker := time.NewTicker(10 * time.Second)
+		defer saveTicker.Stop()
+		for {
+			select {
+			case <-saveTicker.C:
+				// persist in-memory state to disk after every 10secs.
+				logger.LogIf(ctx, ri.updateAfter(ctx, api, 10*time.Second, job))
+
+			case <-ctx.Done():
+				// persist in-memory state immediately before exiting due to context cancellation.
+				logger.LogIf(ctx, ri.updateAfter(ctx, api, 0, job))
+				return
+
+			case <-saverQuitCh:
+				// persist in-memory state immediately to disk.
+				logger.LogIf(ctx, ri.updateAfter(ctx, api, 0, job))
+				return
+			}
+		}
+	}()
+
+	expireCh := make(chan []expireObjInfo, workerSize)
+	go batchObjsForDelete(ctx, r, ri, job, api, wk, expireCh)
+
+	var (
+		prevObj       ObjectInfo
+		matchedFilter BatchJobExpireFilter
+		versionsCount int
+		toDel         []expireObjInfo
+	)
+	for result := range results {
+		// Apply filter to find the matching rule to apply expiry
+		// actions accordingly.
+		// nolint:gocritic
+		if result.IsLatest {
+			// send down filtered entries to be deleted using
+			// DeleteObjects method
+			if len(toDel) > 10 { // batch up to 10 objects/versions to be expired simultaneously.
+				xfer := make([]expireObjInfo, len(toDel))
+				copy(xfer, toDel)
+
+				var done bool
+				select {
+				case <-ctx.Done():
+					done = true
+				case expireCh <- xfer:
+					toDel = toDel[:0] // resetting toDel
+				}
+				if done {
+					break
+				}
+			}
+			var match BatchJobExpireFilter
+			var found bool
+			for _, rule := range r.Rules {
+				if rule.Matches(result, now) {
+					match = rule
+					found = true
+					break
+				}
+			}
+			if !found {
+				continue
+			}
+
+			prevObj = result
+			matchedFilter = match
+			versionsCount = 1
+			// Include the latest version
+			if matchedFilter.Purge.RetainVersions == 0 {
+				toDel = append(toDel, expireObjInfo{
+					ObjectInfo: result,
+					ExpireAll:  true,
+				})
+				continue
+			}
+		} else if prevObj.Name == result.Name {
+			if matchedFilter.Purge.RetainVersions == 0 {
+				continue // including latest version in toDel suffices, skipping other versions
+			}
+			versionsCount++
+		} else {
+			continue
+		}
+
+		if versionsCount <= matchedFilter.Purge.RetainVersions {
+			continue // retain versions
+		}
+		toDel = append(toDel, expireObjInfo{
+			ObjectInfo: result,
+		})
+	}
+	// Send any remaining objects downstream
+	if len(toDel) > 0 {
+		select {
+		case <-ctx.Done():
+		case expireCh <- toDel:
+		}
+	}
+	xioutil.SafeClose(expireCh)
+
+	wk.Wait() // waits for all expire goroutines to complete
+
+	ri.Complete = ri.ObjectsFailed == 0
+	ri.Failed = ri.ObjectsFailed > 0
+	globalBatchJobsMetrics.save(job.ID, ri)
+
+	// Close the saverQuitCh - this also triggers saving in-memory state
+	// immediately one last time before we exit this method.
+	xioutil.SafeClose(saverQuitCh)
+
+	// Notify expire jobs final status to the configured endpoint
+	buf, _ := json.Marshal(ri)
+	if err := r.Notify(context.Background(), bytes.NewReader(buf)); err != nil {
+		logger.LogIf(context.Background(), fmt.Errorf("unable to notify %v", err))
+	}
+
+	return nil
+}
+
+//msgp:ignore batchExpireJobError
+type batchExpireJobError struct {
+	Code           string
+	Description    string
+	HTTPStatusCode int
+}
+
+func (e batchExpireJobError) Error() string {
+	return e.Description
+}
+
+// maxBatchRules maximum number of rules a batch-expiry job supports
+const maxBatchRules = 50
+
+// Validate validates the job definition input
+func (r *BatchJobExpire) Validate(ctx context.Context, job BatchJobRequest, o ObjectLayer) error {
+	if r == nil {
+		return nil
+	}
+
+	if r.APIVersion != batchExpireAPIVersion {
+		return batchExpireJobError{
+			Code:           "InvalidArgument",
+			Description:    "Unsupported batch expire API version",
+			HTTPStatusCode: http.StatusBadRequest,
+		}
+	}
+
+	if r.Bucket == "" {
+		return batchExpireJobError{
+			Code:           "InvalidArgument",
+			Description:    "Bucket argument missing",
+			HTTPStatusCode: http.StatusBadRequest,
+		}
+	}
+
+	if _, err := o.GetBucketInfo(ctx, r.Bucket, BucketOptions{}); err != nil {
+		if isErrBucketNotFound(err) {
+			return batchExpireJobError{
+				Code:           "NoSuchSourceBucket",
+				Description:    "The specified source bucket does not exist",
+				HTTPStatusCode: http.StatusNotFound,
+			}
+		}
+		return err
+	}
+
+	if len(r.Rules) > maxBatchRules {
+		return batchExpireJobError{
+			Code:           "InvalidArgument",
+			Description:    "Too many rules. Batch expire job can't have more than 100 rules",
+			HTTPStatusCode: http.StatusBadRequest,
+		}
+	}
+
+	for _, rule := range r.Rules {
+		if err := rule.Validate(); err != nil {
+			return batchExpireJobError{
+				Code:           "InvalidArgument",
+				Description:    fmt.Sprintf("Invalid batch expire rule: %s", err),
+				HTTPStatusCode: http.StatusBadRequest,
+			}
+		}
+	}
+
+	if err := r.Retry.Validate(); err != nil {
+		return batchExpireJobError{
+			Code:           "InvalidArgument",
+			Description:    fmt.Sprintf("Invalid batch expire retry configuration: %s", err),
+			HTTPStatusCode: http.StatusBadRequest,
+		}
+	}
+	return nil
+}

cmd/batch-expire_gen.go

@@ -0,0 +1,856 @@
+package cmd
+
+// Code generated by github.com/tinylib/msgp DO NOT EDIT.
+
+import (
+	"time"
+
+	"github.com/tinylib/msgp/msgp"
+)
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobExpire) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "APIVersion":
+			z.APIVersion, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "APIVersion")
+				return
+			}
+		case "Bucket":
+			z.Bucket, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Bucket")
+				return
+			}
+		case "Prefix":
+			z.Prefix, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Prefix")
+				return
+			}
+		case "NotificationCfg":
+			err = z.NotificationCfg.DecodeMsg(dc)
+			if err != nil {
+				err = msgp.WrapError(err, "NotificationCfg")
+				return
+			}
+		case "Retry":
+			err = z.Retry.DecodeMsg(dc)
+			if err != nil {
+				err = msgp.WrapError(err, "Retry")
+				return
+			}
+		case "Rules":
+			var zb0002 uint32
+			zb0002, err = dc.ReadArrayHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "Rules")
+				return
+			}
+			if cap(z.Rules) >= int(zb0002) {
+				z.Rules = (z.Rules)[:zb0002]
+			} else {
+				z.Rules = make([]BatchJobExpireFilter, zb0002)
+			}
+			for za0001 := range z.Rules {
+				err = z.Rules[za0001].DecodeMsg(dc)
+				if err != nil {
+					err = msgp.WrapError(err, "Rules", za0001)
+					return
+				}
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *BatchJobExpire) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 6
+	// write "APIVersion"
+	err = en.Append(0x86, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.APIVersion)
+	if err != nil {
+		err = msgp.WrapError(err, "APIVersion")
+		return
+	}
+	// write "Bucket"
+	err = en.Append(0xa6, 0x42, 0x75, 0x63, 0x6b, 0x65, 0x74)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Bucket)
+	if err != nil {
+		err = msgp.WrapError(err, "Bucket")
+		return
+	}
+	// write "Prefix"
+	err = en.Append(0xa6, 0x50, 0x72, 0x65, 0x66, 0x69, 0x78)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Prefix)
+	if err != nil {
+		err = msgp.WrapError(err, "Prefix")
+		return
+	}
+	// write "NotificationCfg"
+	err = en.Append(0xaf, 0x4e, 0x6f, 0x74, 0x69, 0x66, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x66, 0x67)
+	if err != nil {
+		return
+	}
+	err = z.NotificationCfg.EncodeMsg(en)
+	if err != nil {
+		err = msgp.WrapError(err, "NotificationCfg")
+		return
+	}
+	// write "Retry"
+	err = en.Append(0xa5, 0x52, 0x65, 0x74, 0x72, 0x79)
+	if err != nil {
+		return
+	}
+	err = z.Retry.EncodeMsg(en)
+	if err != nil {
+		err = msgp.WrapError(err, "Retry")
+		return
+	}
+	// write "Rules"
+	err = en.Append(0xa5, 0x52, 0x75, 0x6c, 0x65, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteArrayHeader(uint32(len(z.Rules)))
+	if err != nil {
+		err = msgp.WrapError(err, "Rules")
+		return
+	}
+	for za0001 := range z.Rules {
+		err = z.Rules[za0001].EncodeMsg(en)
+		if err != nil {
+			err = msgp.WrapError(err, "Rules", za0001)
+			return
+		}
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *BatchJobExpire) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 6
+	// string "APIVersion"
+	o = append(o, 0x86, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
+	o = msgp.AppendString(o, z.APIVersion)
+	// string "Bucket"
+	o = append(o, 0xa6, 0x42, 0x75, 0x63, 0x6b, 0x65, 0x74)
+	o = msgp.AppendString(o, z.Bucket)
+	// string "Prefix"
+	o = append(o, 0xa6, 0x50, 0x72, 0x65, 0x66, 0x69, 0x78)
+	o = msgp.AppendString(o, z.Prefix)
+	// string "NotificationCfg"
+	o = append(o, 0xaf, 0x4e, 0x6f, 0x74, 0x69, 0x66, 0x69, 0x63, 0x61, 0x74, 0x69, 0x6f, 0x6e, 0x43, 0x66, 0x67)
+	o, err = z.NotificationCfg.MarshalMsg(o)
+	if err != nil {
+		err = msgp.WrapError(err, "NotificationCfg")
+		return
+	}
+	// string "Retry"
+	o = append(o, 0xa5, 0x52, 0x65, 0x74, 0x72, 0x79)
+	o, err = z.Retry.MarshalMsg(o)
+	if err != nil {
+		err = msgp.WrapError(err, "Retry")
+		return
+	}
+	// string "Rules"
+	o = append(o, 0xa5, 0x52, 0x75, 0x6c, 0x65, 0x73)
+	o = msgp.AppendArrayHeader(o, uint32(len(z.Rules)))
+	for za0001 := range z.Rules {
+		o, err = z.Rules[za0001].MarshalMsg(o)
+		if err != nil {
+			err = msgp.WrapError(err, "Rules", za0001)
+			return
+		}
+	}
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobExpire) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "APIVersion":
+			z.APIVersion, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "APIVersion")
+				return
+			}
+		case "Bucket":
+			z.Bucket, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Bucket")
+				return
+			}
+		case "Prefix":
+			z.Prefix, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Prefix")
+				return
+			}
+		case "NotificationCfg":
+			bts, err = z.NotificationCfg.UnmarshalMsg(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "NotificationCfg")
+				return
+			}
+		case "Retry":
+			bts, err = z.Retry.UnmarshalMsg(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Retry")
+				return
+			}
+		case "Rules":
+			var zb0002 uint32
+			zb0002, bts, err = msgp.ReadArrayHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Rules")
+				return
+			}
+			if cap(z.Rules) >= int(zb0002) {
+				z.Rules = (z.Rules)[:zb0002]
+			} else {
+				z.Rules = make([]BatchJobExpireFilter, zb0002)
+			}
+			for za0001 := range z.Rules {
+				bts, err = z.Rules[za0001].UnmarshalMsg(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Rules", za0001)
+					return
+				}
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *BatchJobExpire) Msgsize() (s int) {
+	s = 1 + 11 + msgp.StringPrefixSize + len(z.APIVersion) + 7 + msgp.StringPrefixSize + len(z.Bucket) + 7 + msgp.StringPrefixSize + len(z.Prefix) + 16 + z.NotificationCfg.Msgsize() + 6 + z.Retry.Msgsize() + 6 + msgp.ArrayHeaderSize
+	for za0001 := range z.Rules {
+		s += z.Rules[za0001].Msgsize()
+	}
+	return
+}
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobExpireFilter) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "OlderThan":
+			z.OlderThan, err = dc.ReadDuration()
+			if err != nil {
+				err = msgp.WrapError(err, "OlderThan")
+				return
+			}
+		case "CreatedBefore":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "CreatedBefore")
+					return
+				}
+				z.CreatedBefore = nil
+			} else {
+				if z.CreatedBefore == nil {
+					z.CreatedBefore = new(time.Time)
+				}
+				*z.CreatedBefore, err = dc.ReadTime()
+				if err != nil {
+					err = msgp.WrapError(err, "CreatedBefore")
+					return
+				}
+			}
+		case "Tags":
+			var zb0002 uint32
+			zb0002, err = dc.ReadArrayHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "Tags")
+				return
+			}
+			if cap(z.Tags) >= int(zb0002) {
+				z.Tags = (z.Tags)[:zb0002]
+			} else {
+				z.Tags = make([]BatchJobKV, zb0002)
+			}
+			for za0001 := range z.Tags {
+				err = z.Tags[za0001].DecodeMsg(dc)
+				if err != nil {
+					err = msgp.WrapError(err, "Tags", za0001)
+					return
+				}
+			}
+		case "Metadata":
+			var zb0003 uint32
+			zb0003, err = dc.ReadArrayHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "Metadata")
+				return
+			}
+			if cap(z.Metadata) >= int(zb0003) {
+				z.Metadata = (z.Metadata)[:zb0003]
+			} else {
+				z.Metadata = make([]BatchJobKV, zb0003)
+			}
+			for za0002 := range z.Metadata {
+				err = z.Metadata[za0002].DecodeMsg(dc)
+				if err != nil {
+					err = msgp.WrapError(err, "Metadata", za0002)
+					return
+				}
+			}
+		case "Size":
+			err = z.Size.DecodeMsg(dc)
+			if err != nil {
+				err = msgp.WrapError(err, "Size")
+				return
+			}
+		case "Type":
+			z.Type, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Type")
+				return
+			}
+		case "Name":
+			z.Name, err = dc.ReadString()
+			if err != nil {
+				err = msgp.WrapError(err, "Name")
+				return
+			}
+		case "Purge":
+			var zb0004 uint32
+			zb0004, err = dc.ReadMapHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "Purge")
+				return
+			}
+			for zb0004 > 0 {
+				zb0004--
+				field, err = dc.ReadMapKeyPtr()
+				if err != nil {
+					err = msgp.WrapError(err, "Purge")
+					return
+				}
+				switch msgp.UnsafeString(field) {
+				case "RetainVersions":
+					z.Purge.RetainVersions, err = dc.ReadInt()
+					if err != nil {
+						err = msgp.WrapError(err, "Purge", "RetainVersions")
+						return
+					}
+				default:
+					err = dc.Skip()
+					if err != nil {
+						err = msgp.WrapError(err, "Purge")
+						return
+					}
+				}
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *BatchJobExpireFilter) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 8
+	// write "OlderThan"
+	err = en.Append(0x88, 0xa9, 0x4f, 0x6c, 0x64, 0x65, 0x72, 0x54, 0x68, 0x61, 0x6e)
+	if err != nil {
+		return
+	}
+	err = en.WriteDuration(z.OlderThan)
+	if err != nil {
+		err = msgp.WrapError(err, "OlderThan")
+		return
+	}
+	// write "CreatedBefore"
+	err = en.Append(0xad, 0x43, 0x72, 0x65, 0x61, 0x74, 0x65, 0x64, 0x42, 0x65, 0x66, 0x6f, 0x72, 0x65)
+	if err != nil {
+		return
+	}
+	if z.CreatedBefore == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteTime(*z.CreatedBefore)
+		if err != nil {
+			err = msgp.WrapError(err, "CreatedBefore")
+			return
+		}
+	}
+	// write "Tags"
+	err = en.Append(0xa4, 0x54, 0x61, 0x67, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteArrayHeader(uint32(len(z.Tags)))
+	if err != nil {
+		err = msgp.WrapError(err, "Tags")
+		return
+	}
+	for za0001 := range z.Tags {
+		err = z.Tags[za0001].EncodeMsg(en)
+		if err != nil {
+			err = msgp.WrapError(err, "Tags", za0001)
+			return
+		}
+	}
+	// write "Metadata"
+	err = en.Append(0xa8, 0x4d, 0x65, 0x74, 0x61, 0x64, 0x61, 0x74, 0x61)
+	if err != nil {
+		return
+	}
+	err = en.WriteArrayHeader(uint32(len(z.Metadata)))
+	if err != nil {
+		err = msgp.WrapError(err, "Metadata")
+		return
+	}
+	for za0002 := range z.Metadata {
+		err = z.Metadata[za0002].EncodeMsg(en)
+		if err != nil {
+			err = msgp.WrapError(err, "Metadata", za0002)
+			return
+		}
+	}
+	// write "Size"
+	err = en.Append(0xa4, 0x53, 0x69, 0x7a, 0x65)
+	if err != nil {
+		return
+	}
+	err = z.Size.EncodeMsg(en)
+	if err != nil {
+		err = msgp.WrapError(err, "Size")
+		return
+	}
+	// write "Type"
+	err = en.Append(0xa4, 0x54, 0x79, 0x70, 0x65)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Type)
+	if err != nil {
+		err = msgp.WrapError(err, "Type")
+		return
+	}
+	// write "Name"
+	err = en.Append(0xa4, 0x4e, 0x61, 0x6d, 0x65)
+	if err != nil {
+		return
+	}
+	err = en.WriteString(z.Name)
+	if err != nil {
+		err = msgp.WrapError(err, "Name")
+		return
+	}
+	// write "Purge"
+	err = en.Append(0xa5, 0x50, 0x75, 0x72, 0x67, 0x65)
+	if err != nil {
+		return
+	}
+	// map header, size 1
+	// write "RetainVersions"
+	err = en.Append(0x81, 0xae, 0x52, 0x65, 0x74, 0x61, 0x69, 0x6e, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt(z.Purge.RetainVersions)
+	if err != nil {
+		err = msgp.WrapError(err, "Purge", "RetainVersions")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *BatchJobExpireFilter) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 8
+	// string "OlderThan"
+	o = append(o, 0x88, 0xa9, 0x4f, 0x6c, 0x64, 0x65, 0x72, 0x54, 0x68, 0x61, 0x6e)
+	o = msgp.AppendDuration(o, z.OlderThan)
+	// string "CreatedBefore"
+	o = append(o, 0xad, 0x43, 0x72, 0x65, 0x61, 0x74, 0x65, 0x64, 0x42, 0x65, 0x66, 0x6f, 0x72, 0x65)
+	if z.CreatedBefore == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendTime(o, *z.CreatedBefore)
+	}
+	// string "Tags"
+	o = append(o, 0xa4, 0x54, 0x61, 0x67, 0x73)
+	o = msgp.AppendArrayHeader(o, uint32(len(z.Tags)))
+	for za0001 := range z.Tags {
+		o, err = z.Tags[za0001].MarshalMsg(o)
+		if err != nil {
+			err = msgp.WrapError(err, "Tags", za0001)
+			return
+		}
+	}
+	// string "Metadata"
+	o = append(o, 0xa8, 0x4d, 0x65, 0x74, 0x61, 0x64, 0x61, 0x74, 0x61)
+	o = msgp.AppendArrayHeader(o, uint32(len(z.Metadata)))
+	for za0002 := range z.Metadata {
+		o, err = z.Metadata[za0002].MarshalMsg(o)
+		if err != nil {
+			err = msgp.WrapError(err, "Metadata", za0002)
+			return
+		}
+	}
+	// string "Size"
+	o = append(o, 0xa4, 0x53, 0x69, 0x7a, 0x65)
+	o, err = z.Size.MarshalMsg(o)
+	if err != nil {
+		err = msgp.WrapError(err, "Size")
+		return
+	}
+	// string "Type"
+	o = append(o, 0xa4, 0x54, 0x79, 0x70, 0x65)
+	o = msgp.AppendString(o, z.Type)
+	// string "Name"
+	o = append(o, 0xa4, 0x4e, 0x61, 0x6d, 0x65)
+	o = msgp.AppendString(o, z.Name)
+	// string "Purge"
+	o = append(o, 0xa5, 0x50, 0x75, 0x72, 0x67, 0x65)
+	// map header, size 1
+	// string "RetainVersions"
+	o = append(o, 0x81, 0xae, 0x52, 0x65, 0x74, 0x61, 0x69, 0x6e, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x73)
+	o = msgp.AppendInt(o, z.Purge.RetainVersions)
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobExpireFilter) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "OlderThan":
+			z.OlderThan, bts, err = msgp.ReadDurationBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "OlderThan")
+				return
+			}
+		case "CreatedBefore":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.CreatedBefore = nil
+			} else {
+				if z.CreatedBefore == nil {
+					z.CreatedBefore = new(time.Time)
+				}
+				*z.CreatedBefore, bts, err = msgp.ReadTimeBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "CreatedBefore")
+					return
+				}
+			}
+		case "Tags":
+			var zb0002 uint32
+			zb0002, bts, err = msgp.ReadArrayHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Tags")
+				return
+			}
+			if cap(z.Tags) >= int(zb0002) {
+				z.Tags = (z.Tags)[:zb0002]
+			} else {
+				z.Tags = make([]BatchJobKV, zb0002)
+			}
+			for za0001 := range z.Tags {
+				bts, err = z.Tags[za0001].UnmarshalMsg(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Tags", za0001)
+					return
+				}
+			}
+		case "Metadata":
+			var zb0003 uint32
+			zb0003, bts, err = msgp.ReadArrayHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Metadata")
+				return
+			}
+			if cap(z.Metadata) >= int(zb0003) {
+				z.Metadata = (z.Metadata)[:zb0003]
+			} else {
+				z.Metadata = make([]BatchJobKV, zb0003)
+			}
+			for za0002 := range z.Metadata {
+				bts, err = z.Metadata[za0002].UnmarshalMsg(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Metadata", za0002)
+					return
+				}
+			}
+		case "Size":
+			bts, err = z.Size.UnmarshalMsg(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Size")
+				return
+			}
+		case "Type":
+			z.Type, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Type")
+				return
+			}
+		case "Name":
+			z.Name, bts, err = msgp.ReadStringBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Name")
+				return
+			}
+		case "Purge":
+			var zb0004 uint32
+			zb0004, bts, err = msgp.ReadMapHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Purge")
+				return
+			}
+			for zb0004 > 0 {
+				zb0004--
+				field, bts, err = msgp.ReadMapKeyZC(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Purge")
+					return
+				}
+				switch msgp.UnsafeString(field) {
+				case "RetainVersions":
+					z.Purge.RetainVersions, bts, err = msgp.ReadIntBytes(bts)
+					if err != nil {
+						err = msgp.WrapError(err, "Purge", "RetainVersions")
+						return
+					}
+				default:
+					bts, err = msgp.Skip(bts)
+					if err != nil {
+						err = msgp.WrapError(err, "Purge")
+						return
+					}
+				}
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *BatchJobExpireFilter) Msgsize() (s int) {
+	s = 1 + 10 + msgp.DurationSize + 14
+	if z.CreatedBefore == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.TimeSize
+	}
+	s += 5 + msgp.ArrayHeaderSize
+	for za0001 := range z.Tags {
+		s += z.Tags[za0001].Msgsize()
+	}
+	s += 9 + msgp.ArrayHeaderSize
+	for za0002 := range z.Metadata {
+		s += z.Metadata[za0002].Msgsize()
+	}
+	s += 5 + z.Size.Msgsize() + 5 + msgp.StringPrefixSize + len(z.Type) + 5 + msgp.StringPrefixSize + len(z.Name) + 6 + 1 + 15 + msgp.IntSize
+	return
+}
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobExpirePurge) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "RetainVersions":
+			z.RetainVersions, err = dc.ReadInt()
+			if err != nil {
+				err = msgp.WrapError(err, "RetainVersions")
+				return
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z BatchJobExpirePurge) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 1
+	// write "RetainVersions"
+	err = en.Append(0x81, 0xae, 0x52, 0x65, 0x74, 0x61, 0x69, 0x6e, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt(z.RetainVersions)
+	if err != nil {
+		err = msgp.WrapError(err, "RetainVersions")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z BatchJobExpirePurge) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 1
+	// string "RetainVersions"
+	o = append(o, 0x81, 0xae, 0x52, 0x65, 0x74, 0x61, 0x69, 0x6e, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x73)
+	o = msgp.AppendInt(o, z.RetainVersions)
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobExpirePurge) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "RetainVersions":
+			z.RetainVersions, bts, err = msgp.ReadIntBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "RetainVersions")
+				return
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z BatchJobExpirePurge) Msgsize() (s int) {
+	s = 1 + 15 + msgp.IntSize
+	return
+}

cmd/batch-expire_gen_test.go

@@ -0,0 +1,349 @@
+package cmd
+
+// Code generated by github.com/tinylib/msgp DO NOT EDIT.
+
+import (
+	"bytes"
+	"testing"
+
+	"github.com/tinylib/msgp/msgp"
+)
+
+func TestMarshalUnmarshalBatchJobExpire(t *testing.T) {
+	v := BatchJobExpire{}
+	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 BenchmarkMarshalMsgBatchJobExpire(b *testing.B) {
+	v := BatchJobExpire{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBatchJobExpire(b *testing.B) {
+	v := BatchJobExpire{}
+	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 BenchmarkUnmarshalBatchJobExpire(b *testing.B) {
+	v := BatchJobExpire{}
+	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 TestEncodeDecodeBatchJobExpire(t *testing.T) {
+	v := BatchJobExpire{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBatchJobExpire Msgsize() is inaccurate")
+	}
+
+	vn := BatchJobExpire{}
+	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 BenchmarkEncodeBatchJobExpire(b *testing.B) {
+	v := BatchJobExpire{}
+	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 BenchmarkDecodeBatchJobExpire(b *testing.B) {
+	v := BatchJobExpire{}
+	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 TestMarshalUnmarshalBatchJobExpireFilter(t *testing.T) {
+	v := BatchJobExpireFilter{}
+	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 BenchmarkMarshalMsgBatchJobExpireFilter(b *testing.B) {
+	v := BatchJobExpireFilter{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBatchJobExpireFilter(b *testing.B) {
+	v := BatchJobExpireFilter{}
+	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 BenchmarkUnmarshalBatchJobExpireFilter(b *testing.B) {
+	v := BatchJobExpireFilter{}
+	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 TestEncodeDecodeBatchJobExpireFilter(t *testing.T) {
+	v := BatchJobExpireFilter{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBatchJobExpireFilter Msgsize() is inaccurate")
+	}
+
+	vn := BatchJobExpireFilter{}
+	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 BenchmarkEncodeBatchJobExpireFilter(b *testing.B) {
+	v := BatchJobExpireFilter{}
+	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 BenchmarkDecodeBatchJobExpireFilter(b *testing.B) {
+	v := BatchJobExpireFilter{}
+	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 TestMarshalUnmarshalBatchJobExpirePurge(t *testing.T) {
+	v := BatchJobExpirePurge{}
+	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 BenchmarkMarshalMsgBatchJobExpirePurge(b *testing.B) {
+	v := BatchJobExpirePurge{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBatchJobExpirePurge(b *testing.B) {
+	v := BatchJobExpirePurge{}
+	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 BenchmarkUnmarshalBatchJobExpirePurge(b *testing.B) {
+	v := BatchJobExpirePurge{}
+	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 TestEncodeDecodeBatchJobExpirePurge(t *testing.T) {
+	v := BatchJobExpirePurge{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBatchJobExpirePurge Msgsize() is inaccurate")
+	}
+
+	vn := BatchJobExpirePurge{}
+	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 BenchmarkEncodeBatchJobExpirePurge(b *testing.B) {
+	v := BatchJobExpirePurge{}
+	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 BenchmarkDecodeBatchJobExpirePurge(b *testing.B) {
+	v := BatchJobExpirePurge{}
+	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)
+		}
+	}
+}

cmd/batch-expire_test.go

@@ -0,0 +1,71 @@
+// Copyright (c) 2015-2023 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 (
+	"testing"
+
+	"gopkg.in/yaml.v2"
+)
+
+func TestParseBatchJobExpire(t *testing.T) {
+	expireYaml := `
+expire: # Expire objects that match a condition
+  apiVersion: v1
+  bucket: mybucket # Bucket where this batch job will expire matching objects from
+  prefix: myprefix # (Optional) Prefix under which this job will expire objects matching the rules below.
+  rules:
+    - type: object  # regular objects with zero or more older versions
+      name: NAME # match object names that satisfy the wildcard expression.
+      olderThan: 70h # match objects older than this value
+      createdBefore: "2006-01-02T15:04:05.00Z" # match objects created before "date"
+      tags:
+        - key: name
+          value: pick* # match objects with tag 'name', all values starting with 'pick'
+      metadata:
+        - key: content-type
+          value: image/* # match objects with 'content-type', all values starting with 'image/'
+      size:
+        lessThan: "10MiB" # match objects with size less than this value (e.g. 10MiB)
+        greaterThan: 1MiB # match objects with size greater than this value (e.g. 1MiB)
+      purge:
+          # retainVersions: 0 # (default) delete all versions of the object. This option is the fastest.
+          # retainVersions: 5 # keep the latest 5 versions of the object.
+  
+    - type: deleted # objects with delete marker as their latest version
+      name: NAME # match object names that satisfy the wildcard expression.
+      olderThan: 10h # match objects older than this value (e.g. 7d10h31s)
+      createdBefore: "2006-01-02T15:04:05.00Z" # match objects created before "date"
+      purge:
+          # retainVersions: 0 # (default) delete all versions of the object. This option is the fastest.
+          # retainVersions: 5 # keep the latest 5 versions of the object including delete markers.
+
+  notify:
+    endpoint: https://notify.endpoint # notification endpoint to receive job completion status
+    token: Bearer xxxxx # optional authentication token for the notification endpoint
+  
+  retry:
+    attempts: 10 # number of retries for the job before giving up
+    delay: 500ms # least amount of delay between each retry
+`
+	var job BatchJobRequest
+	err := yaml.UnmarshalStrict([]byte(expireYaml), &job)
+	if err != nil {
+		t.Fatal("Failed to parse batch-job-expire yaml", err)
+	}
+}

cmd/batch-handlers.go

@@ -42,18 +42,22 @@ import (
 	"github.com/minio/minio-go/v7/pkg/credentials"
 	"github.com/minio/minio-go/v7/pkg/encrypt"
 	"github.com/minio/minio-go/v7/pkg/tags"
+	"github.com/minio/minio/internal/config/batch"
 	"github.com/minio/minio/internal/crypto"
 	"github.com/minio/minio/internal/hash"
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/ioutil"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/pkg/v2/console"
 	"github.com/minio/pkg/v2/env"
 	"github.com/minio/pkg/v2/policy"
 	"github.com/minio/pkg/v2/workers"
-	"gopkg.in/yaml.v2"
+	"gopkg.in/yaml.v3"
 )
 
+var globalBatchConfig batch.Config
+
 // BatchJobRequest this is an internal data structure not for external consumption.
 type BatchJobRequest struct {
 	ID        string               `yaml:"-" json:"name"`
@@ -62,6 +66,7 @@ type BatchJobRequest struct {
 	Location  string               `yaml:"-" json:"location"`
 	Replicate *BatchJobReplicateV1 `yaml:"replicate" json:"replicate"`
 	KeyRotate *BatchJobKeyRotateV1 `yaml:"keyrotate" json:"keyrotate"`
+	Expire    *BatchJobExpire      `yaml:"expire" json:"expire"`
 	ctx       context.Context      `msg:"-"`
 }
 
@@ -163,7 +168,7 @@ func (r *BatchJobReplicateV1) ReplicateFromSource(ctx context.Context, api Objec
 	}
 	rd, objInfo, _, err := core.GetObject(ctx, srcBucket, srcObject, gopts)
 	if err != nil {
-		return err
+		return ErrorRespToObjectError(err, srcBucket, srcObject, srcObjInfo.VersionID)
 	}
 	defer rd.Close()
 
@@ -226,7 +231,7 @@ func (r *BatchJobReplicateV1) copyWithMultipartfromSource(ctx context.Context, a
 		}
 		rd, objInfo, _, err := c.GetObject(ctx, srcBucket, srcObject, gopts)
 		if err != nil {
-			return err
+			return ErrorRespToObjectError(err, srcBucket, srcObject, srcObjInfo.VersionID)
 		}
 		defer rd.Close()
 
@@ -262,6 +267,9 @@ func (r *BatchJobReplicateV1) StartFromSource(ctx context.Context, api ObjectLay
 	if err := ri.load(ctx, api, job); err != nil {
 		return err
 	}
+	if ri.Complete {
+		return nil
+	}
 	globalBatchJobsMetrics.save(job.ID, ri)
 
 	delay := job.Replicate.Flags.Retry.Delay
@@ -428,23 +436,27 @@ func (r *BatchJobReplicateV1) StartFromSource(ctx context.Context, api ObjectLay
 			wk.Take()
 			go func() {
 				defer wk.Give()
-				stopFn := globalBatchJobsMetrics.trace(batchReplicationMetricObject, job.ID, attempts, oi)
+				stopFn := globalBatchJobsMetrics.trace(batchJobMetricReplication, job.ID, attempts)
 				success := true
 				if err := r.ReplicateFromSource(ctx, api, core, oi, retry); err != nil {
 					// object must be deleted concurrently, allow these failures but do not count them
 					if isErrVersionNotFound(err) || isErrObjectNotFound(err) {
 						return
 					}
-					stopFn(err)
+					stopFn(oi, err)
 					logger.LogIf(ctx, err)
 					success = false
 				} else {
-					stopFn(nil)
+					stopFn(oi, nil)
 				}
 				ri.trackCurrentBucketObject(r.Target.Bucket, oi, success)
 				globalBatchJobsMetrics.save(job.ID, ri)
 				// persist in-memory state to disk after every 10secs.
 				logger.LogIf(ctx, ri.updateAfter(ctx, api, 10*time.Second, job))
+
+				if wait := globalBatchConfig.ReplicationWait(); wait > 0 {
+					time.Sleep(wait)
+				}
 			}()
 		}
 		wk.Wait()
@@ -517,9 +529,68 @@ func toObjectInfo(bucket, object string, objInfo miniogo.ObjectInfo) ObjectInfo
 		oi.UserDefined[xhttp.AmzStorageClass] = objInfo.StorageClass
 	}
 
+	for k, v := range objInfo.UserMetadata {
+		oi.UserDefined[k] = v
+	}
+
 	return oi
 }
 
+func (r BatchJobReplicateV1) writeAsArchive(ctx context.Context, objAPI ObjectLayer, remoteClnt *minio.Client, entries []ObjectInfo) error {
+	input := make(chan minio.SnowballObject, 1)
+	opts := minio.SnowballOptions{
+		Opts:     minio.PutObjectOptions{},
+		InMemory: *r.Source.Snowball.InMemory,
+		Compress: *r.Source.Snowball.Compress,
+		SkipErrs: *r.Source.Snowball.SkipErrs,
+	}
+
+	go func() {
+		defer xioutil.SafeClose(input)
+
+		for _, entry := range entries {
+			gr, err := objAPI.GetObjectNInfo(ctx, r.Source.Bucket,
+				entry.Name, nil, nil, ObjectOptions{
+					VersionID: entry.VersionID,
+				})
+			if err != nil {
+				logger.LogIf(ctx, err)
+				continue
+			}
+
+			snowballObj := minio.SnowballObject{
+				// Create path to store objects within the bucket.
+				Key:       entry.Name,
+				Size:      entry.Size,
+				ModTime:   entry.ModTime,
+				VersionID: entry.VersionID,
+				Content:   gr,
+				Headers:   make(http.Header),
+				Close: func() {
+					gr.Close()
+				},
+			}
+
+			opts, err := batchReplicationOpts(ctx, "", gr.ObjInfo)
+			if err != nil {
+				logger.LogIf(ctx, err)
+				continue
+			}
+
+			for k, vals := range opts.Header() {
+				for _, v := range vals {
+					snowballObj.Headers.Add(k, v)
+				}
+			}
+
+			input <- snowballObj
+		}
+	}()
+
+	// Collect and upload all entries.
+	return remoteClnt.PutObjectsSnowball(ctx, r.Target.Bucket, opts, input)
+}
+
 // ReplicateToTarget read from source and replicate to configured target
 func (r *BatchJobReplicateV1) ReplicateToTarget(ctx context.Context, api ObjectLayer, c *miniogo.Core, srcObjInfo ObjectInfo, retry bool) error {
 	srcBucket := r.Source.Bucket
@@ -663,7 +734,12 @@ func (ri *batchJobInfo) load(ctx context.Context, api ObjectLayer, job BatchJobR
 		fileName = batchKeyRotationName
 		version = batchKeyRotateVersionV1
 		format = batchKeyRotationFormat
-
+	case job.Expire != nil:
+		fileName = batchExpireName
+		version = batchExpireVersionV1
+		format = batchExpireFormat
+	default:
+		return errors.New("no supported batch job request specified")
 	}
 	data, err := readConfig(ctx, api, pathJoin(job.Location, fileName))
 	if err != nil {
@@ -680,6 +756,11 @@ func (ri *batchJobInfo) load(ctx context.Context, api ObjectLayer, job BatchJobR
 				if job.KeyRotate.Flags.Retry.Attempts > 0 {
 					ri.RetryAttempts = job.KeyRotate.Flags.Retry.Attempts
 				}
+			case job.Expire != nil:
+				ri.RetryAttempts = batchExpireJobDefaultRetries
+				if job.Expire.Retry.Attempts > 0 {
+					ri.RetryAttempts = job.Expire.Retry.Attempts
+				}
 			}
 			return nil
 		}
@@ -789,6 +870,12 @@ func (ri *batchJobInfo) updateAfter(ctx context.Context, api ObjectLayer, durati
 			jobTyp = string(job.Type())
 			fileName = batchKeyRotationName
 			ri.Version = batchKeyRotateVersionV1
+		case madmin.BatchJobExpire:
+			format = batchExpireFormat
+			version = batchExpireVersion
+			jobTyp = string(job.Type())
+			fileName = batchExpireName
+			ri.Version = batchExpireVersionV1
 		default:
 			return errInvalidArgument
 		}
@@ -814,7 +901,18 @@ func (ri *batchJobInfo) updateAfter(ctx context.Context, api ObjectLayer, durati
 	return nil
 }
 
-func (ri *batchJobInfo) trackCurrentBucketObject(bucket string, info ObjectInfo, failed bool) {
+// Note: to be used only with batch jobs that affect multiple versions through
+// a single action. e.g batch-expire has an option to expire all versions of an
+// object which matches the given filters.
+func (ri *batchJobInfo) trackMultipleObjectVersions(bucket string, info ObjectInfo, success bool) {
+	if success {
+		ri.Objects += int64(info.NumVersions)
+	} else {
+		ri.ObjectsFailed += int64(info.NumVersions)
+	}
+}
+
+func (ri *batchJobInfo) trackCurrentBucketObject(bucket string, info ObjectInfo, success bool) {
 	if ri == nil {
 		return
 	}
@@ -824,7 +922,7 @@ func (ri *batchJobInfo) trackCurrentBucketObject(bucket string, info ObjectInfo,
 
 	ri.Bucket = bucket
 	ri.Object = info.Name
-	ri.countItem(info.Size, info.DeleteMarker, failed)
+	ri.countItem(info.Size, info.DeleteMarker, success)
 }
 
 // Start start the batch replication job, resumes if there was a pending job via "job.ID"
@@ -837,6 +935,9 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 	if err := ri.load(ctx, api, job); err != nil {
 		return err
 	}
+	if ri.Complete {
+		return nil
+	}
 	globalBatchJobsMetrics.save(job.ID, ri)
 	lastObject := ri.Object
 
@@ -907,12 +1008,9 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 			// None of the provided metadata filters match skip the object.
 			return false
 		}
-		// if one of source or target is non MinIO, just replicate the top most version like `mc mirror`
-		if (r.Target.Type == BatchJobReplicateResourceS3 || r.Source.Type == BatchJobReplicateResourceS3) && !info.IsLatest {
-			return false
-		}
 
-		return true
+		// if one of source or target is non MinIO, just replicate the top most version like `mc mirror`
+		return !((r.Target.Type == BatchJobReplicateResourceS3 || r.Source.Type == BatchJobReplicateResourceS3) && !info.IsLatest)
 	}
 
 	u, err := url.Parse(r.Target.Endpoint)
@@ -931,8 +1029,73 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 	if err != nil {
 		return err
 	}
+
 	c.SetAppInfo("minio-"+batchJobPrefix, r.APIVersion+" "+job.ID)
 
+	var (
+		walkCh = make(chan ObjectInfo, 100)
+		slowCh = make(chan ObjectInfo, 100)
+	)
+
+	if !*r.Source.Snowball.Disable && r.Source.Type.isMinio() && r.Target.Type.isMinio() {
+		go func() {
+			defer xioutil.SafeClose(slowCh)
+
+			// Snowball currently needs the high level minio-go Client, not the Core one
+			cl, err := miniogo.New(u.Host, &miniogo.Options{
+				Creds:        credentials.NewStaticV4(cred.AccessKey, cred.SecretKey, cred.SessionToken),
+				Secure:       u.Scheme == "https",
+				Transport:    getRemoteInstanceTransport,
+				BucketLookup: lookupStyle(r.Target.Path),
+			})
+			if err != nil {
+				logger.LogIf(ctx, err)
+				return
+			}
+
+			// Already validated before arriving here
+			smallerThan, _ := humanize.ParseBytes(*r.Source.Snowball.SmallerThan)
+
+			var (
+				obj   = ObjectInfo{}
+				batch = make([]ObjectInfo, 0, *r.Source.Snowball.Batch)
+				valid = true
+			)
+
+			for valid {
+				obj, valid = <-walkCh
+
+				if !valid {
+					goto write
+				}
+
+				if obj.DeleteMarker || !obj.VersionPurgeStatus.Empty() || obj.Size >= int64(smallerThan) {
+					slowCh <- obj
+					continue
+				}
+
+				batch = append(batch, obj)
+
+				if len(batch) < *r.Source.Snowball.Batch {
+					continue
+				}
+
+			write:
+				if len(batch) > 0 {
+					if err := r.writeAsArchive(ctx, api, cl, batch); err != nil {
+						logger.LogIf(ctx, err)
+						for _, b := range batch {
+							slowCh <- b
+						}
+					}
+					batch = batch[:0]
+				}
+			}
+		}()
+	} else {
+		slowCh = walkCh
+	}
+
 	workerSize, err := strconv.Atoi(env.Get("_MINIO_BATCH_REPLICATION_WORKERS", strconv.Itoa(runtime.GOMAXPROCS(0)/2)))
 	if err != nil {
 		return err
@@ -944,6 +1107,11 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 		return err
 	}
 
+	walkQuorum := env.Get("_MINIO_BATCH_REPLICATION_WALK_QUORUM", "strict")
+	if walkQuorum == "" {
+		walkQuorum = "strict"
+	}
+
 	retryAttempts := ri.RetryAttempts
 	retry := false
 	for attempts := 1; attempts <= retryAttempts; attempts++ {
@@ -953,10 +1121,10 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 		// one of source/target is s3, skip delete marker and all versions under the same object name.
 		s3Type := r.Target.Type == BatchJobReplicateResourceS3 || r.Source.Type == BatchJobReplicateResourceS3
 
-		results := make(chan ObjectInfo, 100)
-		if err := api.Walk(ctx, r.Source.Bucket, r.Source.Prefix, results, ObjectOptions{
-			WalkMarker: lastObject,
-			WalkFilter: selectObj,
+		if err := api.Walk(ctx, r.Source.Bucket, r.Source.Prefix, walkCh, WalkOptions{
+			Marker:   lastObject,
+			Filter:   selectObj,
+			AskDisks: walkQuorum,
 		}); err != nil {
 			cancel()
 			// Do not need to retry if we can't list objects on source.
@@ -966,7 +1134,7 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 		prevObj := ""
 
 		skipReplicate := false
-		for result := range results {
+		for result := range slowCh {
 			result := result
 			if result.Name != prevObj {
 				prevObj = result.Name
@@ -979,7 +1147,7 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 			go func() {
 				defer wk.Give()
 
-				stopFn := globalBatchJobsMetrics.trace(batchReplicationMetricObject, job.ID, attempts, result)
+				stopFn := globalBatchJobsMetrics.trace(batchJobMetricReplication, job.ID, attempts)
 				success := true
 				if err := r.ReplicateToTarget(ctx, api, c, result, retry); err != nil {
 					if miniogo.ToErrorResponse(err).Code == "PreconditionFailed" {
@@ -990,16 +1158,20 @@ func (r *BatchJobReplicateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 					if isErrVersionNotFound(err) || isErrObjectNotFound(err) {
 						return
 					}
-					stopFn(err)
+					stopFn(result, err)
 					logger.LogIf(ctx, err)
 					success = false
 				} else {
-					stopFn(nil)
+					stopFn(result, nil)
 				}
 				ri.trackCurrentBucketObject(r.Source.Bucket, result, success)
 				globalBatchJobsMetrics.save(job.ID, ri)
 				// persist in-memory state to disk after every 10secs.
 				logger.LogIf(ctx, ri.updateAfter(ctx, api, 10*time.Second, job))
+
+				if wait := globalBatchConfig.ReplicationWait(); wait > 0 {
+					time.Sleep(wait)
+				}
 			}()
 		}
 		wk.Wait()
@@ -1080,6 +1252,9 @@ func (r *BatchJobReplicateV1) Validate(ctx context.Context, job BatchJobRequest,
 	if err := r.Source.Type.Validate(); err != nil {
 		return err
 	}
+	if err := r.Source.Snowball.Validate(); err != nil {
+		return err
+	}
 	if r.Source.Creds.Empty() && r.Target.Creds.Empty() {
 		return errInvalidArgument
 	}
@@ -1201,6 +1376,8 @@ func (j BatchJobRequest) Type() madmin.BatchJobType {
 		return madmin.BatchJobReplicate
 	case j.KeyRotate != nil:
 		return madmin.BatchJobKeyRotate
+	case j.Expire != nil:
+		return madmin.BatchJobExpire
 	}
 	return madmin.BatchJobType("unknown")
 }
@@ -1213,6 +1390,8 @@ func (j BatchJobRequest) Validate(ctx context.Context, o ObjectLayer) error {
 		return j.Replicate.Validate(ctx, j, o)
 	case j.KeyRotate != nil:
 		return j.KeyRotate.Validate(ctx, j, o)
+	case j.Expire != nil:
+		return j.Expire.Validate(ctx, j, o)
 	}
 	return errInvalidArgument
 }
@@ -1223,12 +1402,14 @@ func (j BatchJobRequest) delete(ctx context.Context, api ObjectLayer) {
 		deleteConfig(ctx, api, pathJoin(j.Location, batchReplName))
 	case j.KeyRotate != nil:
 		deleteConfig(ctx, api, pathJoin(j.Location, batchKeyRotationName))
+	case j.Expire != nil:
+		deleteConfig(ctx, api, pathJoin(j.Location, batchExpireName))
 	}
 	deleteConfig(ctx, api, j.Location)
 }
 
 func (j *BatchJobRequest) save(ctx context.Context, api ObjectLayer) error {
-	if j.Replicate == nil && j.KeyRotate == nil {
+	if j.Replicate == nil && j.KeyRotate == nil && j.Expire == nil {
 		return errInvalidArgument
 	}
 
@@ -1297,7 +1478,7 @@ func (a adminAPIHandlers) ListBatchJobs(w http.ResponseWriter, r *http.Request)
 	ctx, cancel := context.WithCancel(ctx)
 	defer cancel()
 
-	if err := objectAPI.Walk(ctx, minioMetaBucket, batchJobPrefix, resultCh, ObjectOptions{}); err != nil {
+	if err := objectAPI.Walk(ctx, minioMetaBucket, batchJobPrefix, resultCh, WalkOptions{}); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1384,7 +1565,35 @@ func (a adminAPIHandlers) StartBatchJob(w http.ResponseWriter, r *http.Request)
 	}
 
 	job := &BatchJobRequest{}
-	if err = yaml.UnmarshalStrict(buf, job); err != nil {
+	if err = yaml.Unmarshal(buf, job); err != nil {
+		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+
+	// Fill with default values
+	if job.Replicate != nil {
+		if job.Replicate.Source.Snowball.Disable == nil {
+			job.Replicate.Source.Snowball.Disable = ptr(false)
+		}
+		if job.Replicate.Source.Snowball.Batch == nil {
+			job.Replicate.Source.Snowball.Batch = ptr(100)
+		}
+		if job.Replicate.Source.Snowball.InMemory == nil {
+			job.Replicate.Source.Snowball.InMemory = ptr(true)
+		}
+		if job.Replicate.Source.Snowball.Compress == nil {
+			job.Replicate.Source.Snowball.Compress = ptr(false)
+		}
+		if job.Replicate.Source.Snowball.SmallerThan == nil {
+			job.Replicate.Source.Snowball.SmallerThan = ptr("5MiB")
+		}
+		if job.Replicate.Source.Snowball.SkipErrs == nil {
+			job.Replicate.Source.Snowball.SkipErrs = ptr(true)
+		}
+	}
+
+	//  Validate the incoming job request
+	if err := job.Validate(ctx, objectAPI); err != nil {
 		writeErrorResponseJSON(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1486,7 +1695,7 @@ func (j *BatchJobPool) resume() {
 	results := make(chan ObjectInfo, 100)
 	ctx, cancel := context.WithCancel(j.ctx)
 	defer cancel()
-	if err := j.objLayer.Walk(ctx, minioMetaBucket, batchJobPrefix, results, ObjectOptions{}); err != nil {
+	if err := j.objLayer.Walk(ctx, minioMetaBucket, batchJobPrefix, results, WalkOptions{}); err != nil {
 		logger.LogIf(j.ctx, err)
 		return
 	}
@@ -1525,7 +1734,8 @@ func (j *BatchJobPool) AddWorker() {
 			if !ok {
 				return
 			}
-			if job.Replicate != nil {
+			switch {
+			case job.Replicate != nil:
 				if job.Replicate.RemoteToLocal() {
 					if err := job.Replicate.StartFromSource(job.ctx, j.objLayer, *job); err != nil {
 						if !isErrBucketNotFound(err) {
@@ -1545,14 +1755,20 @@ func (j *BatchJobPool) AddWorker() {
 						// Bucket not found proceed to delete such a job.
 					}
 				}
-			}
-			if job.KeyRotate != nil {
+			case job.KeyRotate != nil:
 				if err := job.KeyRotate.Start(job.ctx, j.objLayer, *job); err != nil {
 					if !isErrBucketNotFound(err) {
 						logger.LogIf(j.ctx, err)
 						continue
 					}
 				}
+			case job.Expire != nil:
+				if err := job.Expire.Start(job.ctx, j.objLayer, *job); err != nil {
+					if !isErrBucketNotFound(err) {
+						logger.LogIf(j.ctx, err)
+						continue
+					}
+				}
 			}
 			job.delete(j.ctx, j.objLayer)
 			j.canceler(job.ID, false)
@@ -1594,7 +1810,7 @@ func (j *BatchJobPool) queueJob(req *BatchJobRequest) error {
 	select {
 	case <-j.ctx.Done():
 		j.once.Do(func() {
-			close(j.jobCh)
+			xioutil.SafeClose(j.jobCh)
 		})
 	case j.jobCh <- req:
 	default:
@@ -1630,24 +1846,26 @@ type batchJobMetrics struct {
 type batchJobMetric uint8
 
 const (
-	batchReplicationMetricObject batchJobMetric = iota
-	batchKeyRotationMetricObject
+	batchJobMetricReplication batchJobMetric = iota
+	batchJobMetricKeyRotation
+	batchJobMetricExpire
 )
 
-func batchJobTrace(d batchJobMetric, job string, startTime time.Time, duration time.Duration, info ObjectInfo, attempts int, err error) madmin.TraceInfo {
+func batchJobTrace(d batchJobMetric, job string, startTime time.Time, duration time.Duration, info objTraceInfoer, attempts int, err error) madmin.TraceInfo {
 	var errStr string
 	if err != nil {
 		errStr = err.Error()
 	}
-	jobKind := "batchReplication"
 	traceType := madmin.TraceBatchReplication
-	if d == batchKeyRotationMetricObject {
-		jobKind = "batchKeyRotation"
+	switch d {
+	case batchJobMetricKeyRotation:
 		traceType = madmin.TraceBatchKeyRotation
+	case batchJobMetricExpire:
+		traceType = madmin.TraceBatchExpire
 	}
-	funcName := fmt.Sprintf("%s.%s (job-name=%s)", jobKind, d.String(), job)
+	funcName := fmt.Sprintf("%s() (job-name=%s)", d.String(), job)
 	if attempts > 0 {
-		funcName = fmt.Sprintf("%s.%s (job-name=%s,attempts=%s)", jobKind, d.String(), job, humanize.Ordinal(attempts))
+		funcName = fmt.Sprintf("%s() (job-name=%s,attempts=%s)", d.String(), job, humanize.Ordinal(attempts))
 	}
 	return madmin.TraceInfo{
 		TraceType: traceType,
@@ -1655,55 +1873,65 @@ func batchJobTrace(d batchJobMetric, job string, startTime time.Time, duration t
 		NodeName:  globalLocalNodeName,
 		FuncName:  funcName,
 		Duration:  duration,
-		Path:      info.Name,
+		Path:      fmt.Sprintf("%s (versionID=%s)", info.TraceObjName(), info.TraceVersionID()),
 		Error:     errStr,
 	}
 }
 
+func (ri *batchJobInfo) metric() madmin.JobMetric {
+	m := madmin.JobMetric{
+		JobID:         ri.JobID,
+		JobType:       ri.JobType,
+		StartTime:     ri.StartTime,
+		LastUpdate:    ri.LastUpdate,
+		RetryAttempts: ri.RetryAttempts,
+		Complete:      ri.Complete,
+		Failed:        ri.Failed,
+	}
+
+	switch ri.JobType {
+	case string(madmin.BatchJobReplicate):
+		m.Replicate = &madmin.ReplicateInfo{
+			Bucket:           ri.Bucket,
+			Object:           ri.Object,
+			Objects:          ri.Objects,
+			ObjectsFailed:    ri.ObjectsFailed,
+			BytesTransferred: ri.BytesTransferred,
+			BytesFailed:      ri.BytesFailed,
+		}
+	case string(madmin.BatchJobKeyRotate):
+		m.KeyRotate = &madmin.KeyRotationInfo{
+			Bucket:        ri.Bucket,
+			Object:        ri.Object,
+			Objects:       ri.Objects,
+			ObjectsFailed: ri.ObjectsFailed,
+		}
+	case string(madmin.BatchJobExpire):
+		m.Expired = &madmin.ExpirationInfo{
+			Bucket:        ri.Bucket,
+			Object:        ri.Object,
+			Objects:       ri.Objects,
+			ObjectsFailed: ri.ObjectsFailed,
+		}
+	}
+
+	return m
+}
+
 func (m *batchJobMetrics) report(jobID string) (metrics *madmin.BatchJobMetrics) {
 	metrics = &madmin.BatchJobMetrics{CollectedAt: time.Now(), Jobs: make(map[string]madmin.JobMetric)}
 	m.RLock()
 	defer m.RUnlock()
 
-	match := true
+	if jobID != "" {
+		if job, ok := m.metrics[jobID]; ok {
+			metrics.Jobs[jobID] = job.metric()
+		}
+		return metrics
+	}
+
 	for id, job := range m.metrics {
-		if jobID != "" {
-			match = id == jobID
-		}
-		if !match {
-			continue
-		}
-
-		m := madmin.JobMetric{
-			JobID:         job.JobID,
-			JobType:       job.JobType,
-			StartTime:     job.StartTime,
-			LastUpdate:    job.LastUpdate,
-			RetryAttempts: job.RetryAttempts,
-			Complete:      job.Complete,
-			Failed:        job.Failed,
-		}
-
-		switch job.JobType {
-		case string(madmin.BatchJobReplicate):
-			m.Replicate = &madmin.ReplicateInfo{
-				Bucket:           job.Bucket,
-				Object:           job.Object,
-				Objects:          job.Objects,
-				ObjectsFailed:    job.ObjectsFailed,
-				BytesTransferred: job.BytesTransferred,
-				BytesFailed:      job.BytesFailed,
-			}
-		case string(madmin.BatchJobKeyRotate):
-			m.KeyRotate = &madmin.KeyRotationInfo{
-				Bucket:        job.Bucket,
-				Object:        job.Object,
-				Objects:       job.Objects,
-				ObjectsFailed: job.ObjectsFailed,
-			}
-		}
-
-		metrics.Jobs[id] = m
+		metrics.Jobs[id] = job.metric()
 	}
 	return metrics
 }
@@ -1748,19 +1976,52 @@ func (m *batchJobMetrics) save(jobID string, ri *batchJobInfo) {
 	m.metrics[jobID] = ri.clone()
 }
 
-func (m *batchJobMetrics) trace(d batchJobMetric, job string, attempts int, info ObjectInfo) func(err error) {
+type objTraceInfoer interface {
+	TraceObjName() string
+	TraceVersionID() string
+}
+
+// TraceObjName returns name of object being traced
+func (td ObjectToDelete) TraceObjName() string {
+	return td.ObjectName
+}
+
+// TraceVersionID returns version-id of object being traced
+func (td ObjectToDelete) TraceVersionID() string {
+	return td.VersionID
+}
+
+// TraceObjName returns name of object being traced
+func (oi ObjectInfo) TraceObjName() string {
+	return oi.Name
+}
+
+// TraceVersionID returns version-id of object being traced
+func (oi ObjectInfo) TraceVersionID() string {
+	return oi.VersionID
+}
+
+func (m *batchJobMetrics) trace(d batchJobMetric, job string, attempts int) func(info objTraceInfoer, err error) {
 	startTime := time.Now()
-	return func(err error) {
+	return func(info objTraceInfoer, err error) {
 		duration := time.Since(startTime)
+		if globalTrace.NumSubscribers(madmin.TraceBatch) > 0 {
+			globalTrace.Publish(batchJobTrace(d, job, startTime, duration, info, attempts, err))
+			return
+		}
 		switch d {
-		case batchReplicationMetricObject:
+		case batchJobMetricReplication:
 			if globalTrace.NumSubscribers(madmin.TraceBatchReplication) > 0 {
 				globalTrace.Publish(batchJobTrace(d, job, startTime, duration, info, attempts, err))
 			}
-		case batchKeyRotationMetricObject:
+		case batchJobMetricKeyRotation:
 			if globalTrace.NumSubscribers(madmin.TraceBatchKeyRotation) > 0 {
 				globalTrace.Publish(batchJobTrace(d, job, startTime, duration, info, attempts, err))
 			}
+		case batchJobMetricExpire:
+			if globalTrace.NumSubscribers(madmin.TraceBatchExpire) > 0 {
+				globalTrace.Publish(batchJobTrace(d, job, startTime, duration, info, attempts, err))
+			}
 		}
 	}
 }

cmd/batch-handlers_gen.go

@@ -84,6 +84,24 @@ func (z *BatchJobRequest) DecodeMsg(dc *msgp.Reader) (err error) {
 					return
 				}
 			}
+		case "Expire":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "Expire")
+					return
+				}
+				z.Expire = nil
+			} else {
+				if z.Expire == nil {
+					z.Expire = new(BatchJobExpire)
+				}
+				err = z.Expire.DecodeMsg(dc)
+				if err != nil {
+					err = msgp.WrapError(err, "Expire")
+					return
+				}
+			}
 		default:
 			err = dc.Skip()
 			if err != nil {
@@ -97,9 +115,9 @@ func (z *BatchJobRequest) DecodeMsg(dc *msgp.Reader) (err error) {
 
 // EncodeMsg implements msgp.Encodable
 func (z *BatchJobRequest) EncodeMsg(en *msgp.Writer) (err error) {
-	// map header, size 6
+	// map header, size 7
 	// write "ID"
-	err = en.Append(0x86, 0xa2, 0x49, 0x44)
+	err = en.Append(0x87, 0xa2, 0x49, 0x44)
 	if err != nil {
 		return
 	}
@@ -172,15 +190,32 @@ func (z *BatchJobRequest) EncodeMsg(en *msgp.Writer) (err error) {
 			return
 		}
 	}
+	// write "Expire"
+	err = en.Append(0xa6, 0x45, 0x78, 0x70, 0x69, 0x72, 0x65)
+	if err != nil {
+		return
+	}
+	if z.Expire == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = z.Expire.EncodeMsg(en)
+		if err != nil {
+			err = msgp.WrapError(err, "Expire")
+			return
+		}
+	}
 	return
 }
 
 // MarshalMsg implements msgp.Marshaler
 func (z *BatchJobRequest) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.Require(b, z.Msgsize())
-	// map header, size 6
+	// map header, size 7
 	// string "ID"
-	o = append(o, 0x86, 0xa2, 0x49, 0x44)
+	o = append(o, 0x87, 0xa2, 0x49, 0x44)
 	o = msgp.AppendString(o, z.ID)
 	// string "User"
 	o = append(o, 0xa4, 0x55, 0x73, 0x65, 0x72)
@@ -213,6 +248,17 @@ func (z *BatchJobRequest) MarshalMsg(b []byte) (o []byte, err error) {
 			return
 		}
 	}
+	// string "Expire"
+	o = append(o, 0xa6, 0x45, 0x78, 0x70, 0x69, 0x72, 0x65)
+	if z.Expire == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o, err = z.Expire.MarshalMsg(o)
+		if err != nil {
+			err = msgp.WrapError(err, "Expire")
+			return
+		}
+	}
 	return
 }
 
@@ -292,6 +338,23 @@ func (z *BatchJobRequest) UnmarshalMsg(bts []byte) (o []byte, err error) {
 					return
 				}
 			}
+		case "Expire":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.Expire = nil
+			} else {
+				if z.Expire == nil {
+					z.Expire = new(BatchJobExpire)
+				}
+				bts, err = z.Expire.UnmarshalMsg(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Expire")
+					return
+				}
+			}
 		default:
 			bts, err = msgp.Skip(bts)
 			if err != nil {
@@ -318,6 +381,12 @@ func (z *BatchJobRequest) Msgsize() (s int) {
 	} else {
 		s += z.KeyRotate.Msgsize()
 	}
+	s += 7
+	if z.Expire == nil {
+		s += msgp.NilSize
+	} else {
+		s += z.Expire.Msgsize()
+	}
 	return
 }
 

cmd/batch-job-common-types.go

@@ -18,24 +18,66 @@
 package cmd
 
 import (
+	"fmt"
 	"strings"
 	"time"
 
+	"github.com/dustin/go-humanize"
 	"github.com/minio/pkg/v2/wildcard"
+	"gopkg.in/yaml.v3"
 )
 
 //go:generate msgp -file $GOFILE
+//msgp:ignore BatchJobYamlErr
+
+// BatchJobYamlErr can be used to return yaml validation errors with line,
+// column information guiding user to fix syntax errors
+type BatchJobYamlErr struct {
+	line, col int
+	msg       string
+}
+
+// message returns the error message excluding line, col information.
+// Intended to be used in unit tests.
+func (b BatchJobYamlErr) message() string {
+	return b.msg
+}
+
+// Error implements Error interface
+func (b BatchJobYamlErr) Error() string {
+	return fmt.Sprintf("%s\n Hint: error near line: %d, col: %d", b.msg, b.line, b.col)
+}
 
 // BatchJobKV is a key-value data type which supports wildcard matching
 type BatchJobKV struct {
-	Key   string `yaml:"key" json:"key"`
-	Value string `yaml:"value" json:"value"`
+	line, col int
+	Key       string `yaml:"key" json:"key"`
+	Value     string `yaml:"value" json:"value"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobKV{}
+
+// UnmarshalYAML - BatchJobKV extends default unmarshal to extract line, col information.
+func (kv *BatchJobKV) UnmarshalYAML(val *yaml.Node) error {
+	type jobKV BatchJobKV
+	var tmp jobKV
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+	*kv = BatchJobKV(tmp)
+	kv.line, kv.col = val.Line, val.Column
+	return nil
 }
 
 // Validate returns an error if key is empty
 func (kv BatchJobKV) Validate() error {
 	if kv.Key == "" {
-		return errInvalidArgument
+		return BatchJobYamlErr{
+			line: kv.line,
+			col:  kv.col,
+			msg:  "key can't be empty",
+		}
 	}
 	return nil
 }
@@ -59,25 +101,190 @@ func (kv BatchJobKV) Match(ikv BatchJobKV) bool {
 // BatchJobNotification stores notification endpoint and token information.
 // Used by batch jobs to notify of their status.
 type BatchJobNotification struct {
-	Endpoint string `yaml:"endpoint" json:"endpoint"`
-	Token    string `yaml:"token" json:"token"`
+	line, col int
+	Endpoint  string `yaml:"endpoint" json:"endpoint"`
+	Token     string `yaml:"token" json:"token"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobNotification{}
+
+// UnmarshalYAML - BatchJobNotification extends unmarshal to extract line, column information
+func (b *BatchJobNotification) UnmarshalYAML(val *yaml.Node) error {
+	type notification BatchJobNotification
+	var tmp notification
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*b = BatchJobNotification(tmp)
+	b.line, b.col = val.Line, val.Column
+	return nil
 }
 
 // BatchJobRetry stores retry configuration used in the event of failures.
 type BatchJobRetry struct {
-	Attempts int           `yaml:"attempts" json:"attempts"` // number of retry attempts
-	Delay    time.Duration `yaml:"delay" json:"delay"`       // delay between each retries
+	line, col int
+	Attempts  int           `yaml:"attempts" json:"attempts"` // number of retry attempts
+	Delay     time.Duration `yaml:"delay" json:"delay"`       // delay between each retries
+}
+
+var _ yaml.Unmarshaler = &BatchJobRetry{}
+
+// UnmarshalYAML - BatchJobRetry extends unmarshal to extract line, column information
+func (r *BatchJobRetry) UnmarshalYAML(val *yaml.Node) error {
+	type retry BatchJobRetry
+	var tmp retry
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*r = BatchJobRetry(tmp)
+	r.line, r.col = val.Line, val.Column
+	return nil
 }
 
 // Validate validates input replicate retries.
 func (r BatchJobRetry) Validate() error {
 	if r.Attempts < 0 {
-		return errInvalidArgument
+		return BatchJobYamlErr{
+			line: r.line,
+			col:  r.col,
+			msg:  "Invalid arguments specified",
+		}
 	}
 
 	if r.Delay < 0 {
-		return errInvalidArgument
+		return BatchJobYamlErr{
+			line: r.line,
+			col:  r.col,
+			msg:  "Invalid arguments specified",
+		}
 	}
 
 	return nil
 }
+
+//   # snowball based archive transfer is by default enabled when source
+//   # is local and target is remote which is also minio.
+//   snowball:
+//     disable: false # optionally turn-off snowball archive transfer
+//     batch: 100 # upto this many objects per archive
+//     inmemory: true # indicates if the archive must be staged locally or in-memory
+//     compress: true # S2/Snappy compressed archive
+//     smallerThan: 5MiB # create archive for all objects smaller than 5MiB
+//     skipErrs: false # skips any source side read() errors
+
+// BatchJobSnowball describes the snowball feature when replicating objects from a local source to a remote target
+type BatchJobSnowball struct {
+	line, col   int
+	Disable     *bool   `yaml:"disable" json:"disable"`
+	Batch       *int    `yaml:"batch" json:"batch"`
+	InMemory    *bool   `yaml:"inmemory" json:"inmemory"`
+	Compress    *bool   `yaml:"compress" json:"compress"`
+	SmallerThan *string `yaml:"smallerThan" json:"smallerThan"`
+	SkipErrs    *bool   `yaml:"skipErrs" json:"skipErrs"`
+}
+
+var _ yaml.Unmarshaler = &BatchJobSnowball{}
+
+// UnmarshalYAML - BatchJobSnowball extends unmarshal to extract line, column information
+func (b *BatchJobSnowball) UnmarshalYAML(val *yaml.Node) error {
+	type snowball BatchJobSnowball
+	var tmp snowball
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*b = BatchJobSnowball(tmp)
+	b.line, b.col = val.Line, val.Column
+	return nil
+}
+
+// Validate the snowball parameters in the job description
+func (b BatchJobSnowball) Validate() error {
+	if *b.Batch <= 0 {
+		return BatchJobYamlErr{
+			line: b.line,
+			col:  b.col,
+			msg:  "batch number should be non positive zero",
+		}
+	}
+	_, err := humanize.ParseBytes(*b.SmallerThan)
+	if err != nil {
+		return BatchJobYamlErr{
+			line: b.line,
+			col:  b.col,
+			msg:  err.Error(),
+		}
+	}
+	return nil
+}
+
+// BatchJobSizeFilter supports size based filters - LesserThan and GreaterThan
+type BatchJobSizeFilter struct {
+	line, col  int
+	UpperBound BatchJobSize `yaml:"lessThan" json:"lessThan"`
+	LowerBound BatchJobSize `yaml:"greaterThan" json:"greaterThan"`
+}
+
+// UnmarshalYAML - BatchJobSizeFilter extends unmarshal to extract line, column information
+func (sf *BatchJobSizeFilter) UnmarshalYAML(val *yaml.Node) error {
+	type sizeFilter BatchJobSizeFilter
+	var tmp sizeFilter
+	err := val.Decode(&tmp)
+	if err != nil {
+		return err
+	}
+
+	*sf = BatchJobSizeFilter(tmp)
+	sf.line, sf.col = val.Line, val.Column
+	return nil
+}
+
+// InRange returns true in the following cases and false otherwise,
+// - sf.LowerBound < sz, when sf.LowerBound alone is specified
+// - sz < sf.UpperBound, when sf.UpperBound alone is specified
+// - sf.LowerBound < sz < sf.UpperBound when both are specified,
+func (sf BatchJobSizeFilter) InRange(sz int64) bool {
+	if sf.UpperBound > 0 && sz > int64(sf.UpperBound) {
+		return false
+	}
+
+	if sf.LowerBound > 0 && sz < int64(sf.LowerBound) {
+		return false
+	}
+	return true
+}
+
+// Validate checks if sf is a valid batch-job size filter
+func (sf BatchJobSizeFilter) Validate() error {
+	if sf.LowerBound > 0 && sf.UpperBound > 0 && sf.LowerBound >= sf.UpperBound {
+		return BatchJobYamlErr{
+			line: sf.line,
+			col:  sf.col,
+			msg:  "invalid batch-job size filter",
+		}
+	}
+	return nil
+}
+
+// BatchJobSize supports humanized byte values in yaml files type BatchJobSize uint64
+type BatchJobSize int64
+
+// UnmarshalYAML to parse humanized byte values
+func (s *BatchJobSize) UnmarshalYAML(unmarshal func(interface{}) error) error {
+	var batchExpireSz string
+	err := unmarshal(&batchExpireSz)
+	if err != nil {
+		return err
+	}
+	sz, err := humanize.ParseBytes(batchExpireSz)
+	if err != nil {
+		return err
+	}
+	*s = BatchJobSize(sz)
+	return nil
+}

cmd/batch-job-common-types_gen.go

@@ -389,3 +389,666 @@ func (z BatchJobRetry) Msgsize() (s int) {
 	s = 1 + 9 + msgp.IntSize + 6 + msgp.DurationSize
 	return
 }
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobSize) DecodeMsg(dc *msgp.Reader) (err error) {
+	{
+		var zb0001 int64
+		zb0001, err = dc.ReadInt64()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		(*z) = BatchJobSize(zb0001)
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z BatchJobSize) EncodeMsg(en *msgp.Writer) (err error) {
+	err = en.WriteInt64(int64(z))
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z BatchJobSize) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	o = msgp.AppendInt64(o, int64(z))
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobSize) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	{
+		var zb0001 int64
+		zb0001, bts, err = msgp.ReadInt64Bytes(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		(*z) = BatchJobSize(zb0001)
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z BatchJobSize) Msgsize() (s int) {
+	s = msgp.Int64Size
+	return
+}
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobSizeFilter) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "UpperBound":
+			{
+				var zb0002 int64
+				zb0002, err = dc.ReadInt64()
+				if err != nil {
+					err = msgp.WrapError(err, "UpperBound")
+					return
+				}
+				z.UpperBound = BatchJobSize(zb0002)
+			}
+		case "LowerBound":
+			{
+				var zb0003 int64
+				zb0003, err = dc.ReadInt64()
+				if err != nil {
+					err = msgp.WrapError(err, "LowerBound")
+					return
+				}
+				z.LowerBound = BatchJobSize(zb0003)
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z BatchJobSizeFilter) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 2
+	// write "UpperBound"
+	err = en.Append(0x82, 0xaa, 0x55, 0x70, 0x70, 0x65, 0x72, 0x42, 0x6f, 0x75, 0x6e, 0x64)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt64(int64(z.UpperBound))
+	if err != nil {
+		err = msgp.WrapError(err, "UpperBound")
+		return
+	}
+	// write "LowerBound"
+	err = en.Append(0xaa, 0x4c, 0x6f, 0x77, 0x65, 0x72, 0x42, 0x6f, 0x75, 0x6e, 0x64)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt64(int64(z.LowerBound))
+	if err != nil {
+		err = msgp.WrapError(err, "LowerBound")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z BatchJobSizeFilter) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 2
+	// string "UpperBound"
+	o = append(o, 0x82, 0xaa, 0x55, 0x70, 0x70, 0x65, 0x72, 0x42, 0x6f, 0x75, 0x6e, 0x64)
+	o = msgp.AppendInt64(o, int64(z.UpperBound))
+	// string "LowerBound"
+	o = append(o, 0xaa, 0x4c, 0x6f, 0x77, 0x65, 0x72, 0x42, 0x6f, 0x75, 0x6e, 0x64)
+	o = msgp.AppendInt64(o, int64(z.LowerBound))
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobSizeFilter) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "UpperBound":
+			{
+				var zb0002 int64
+				zb0002, bts, err = msgp.ReadInt64Bytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "UpperBound")
+					return
+				}
+				z.UpperBound = BatchJobSize(zb0002)
+			}
+		case "LowerBound":
+			{
+				var zb0003 int64
+				zb0003, bts, err = msgp.ReadInt64Bytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "LowerBound")
+					return
+				}
+				z.LowerBound = BatchJobSize(zb0003)
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z BatchJobSizeFilter) Msgsize() (s int) {
+	s = 1 + 11 + msgp.Int64Size + 11 + msgp.Int64Size
+	return
+}
+
+// DecodeMsg implements msgp.Decodable
+func (z *BatchJobSnowball) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "Disable":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "Disable")
+					return
+				}
+				z.Disable = nil
+			} else {
+				if z.Disable == nil {
+					z.Disable = new(bool)
+				}
+				*z.Disable, err = dc.ReadBool()
+				if err != nil {
+					err = msgp.WrapError(err, "Disable")
+					return
+				}
+			}
+		case "Batch":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "Batch")
+					return
+				}
+				z.Batch = nil
+			} else {
+				if z.Batch == nil {
+					z.Batch = new(int)
+				}
+				*z.Batch, err = dc.ReadInt()
+				if err != nil {
+					err = msgp.WrapError(err, "Batch")
+					return
+				}
+			}
+		case "InMemory":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "InMemory")
+					return
+				}
+				z.InMemory = nil
+			} else {
+				if z.InMemory == nil {
+					z.InMemory = new(bool)
+				}
+				*z.InMemory, err = dc.ReadBool()
+				if err != nil {
+					err = msgp.WrapError(err, "InMemory")
+					return
+				}
+			}
+		case "Compress":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "Compress")
+					return
+				}
+				z.Compress = nil
+			} else {
+				if z.Compress == nil {
+					z.Compress = new(bool)
+				}
+				*z.Compress, err = dc.ReadBool()
+				if err != nil {
+					err = msgp.WrapError(err, "Compress")
+					return
+				}
+			}
+		case "SmallerThan":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "SmallerThan")
+					return
+				}
+				z.SmallerThan = nil
+			} else {
+				if z.SmallerThan == nil {
+					z.SmallerThan = new(string)
+				}
+				*z.SmallerThan, err = dc.ReadString()
+				if err != nil {
+					err = msgp.WrapError(err, "SmallerThan")
+					return
+				}
+			}
+		case "SkipErrs":
+			if dc.IsNil() {
+				err = dc.ReadNil()
+				if err != nil {
+					err = msgp.WrapError(err, "SkipErrs")
+					return
+				}
+				z.SkipErrs = nil
+			} else {
+				if z.SkipErrs == nil {
+					z.SkipErrs = new(bool)
+				}
+				*z.SkipErrs, err = dc.ReadBool()
+				if err != nil {
+					err = msgp.WrapError(err, "SkipErrs")
+					return
+				}
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *BatchJobSnowball) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 6
+	// write "Disable"
+	err = en.Append(0x86, 0xa7, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c, 0x65)
+	if err != nil {
+		return
+	}
+	if z.Disable == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteBool(*z.Disable)
+		if err != nil {
+			err = msgp.WrapError(err, "Disable")
+			return
+		}
+	}
+	// write "Batch"
+	err = en.Append(0xa5, 0x42, 0x61, 0x74, 0x63, 0x68)
+	if err != nil {
+		return
+	}
+	if z.Batch == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteInt(*z.Batch)
+		if err != nil {
+			err = msgp.WrapError(err, "Batch")
+			return
+		}
+	}
+	// write "InMemory"
+	err = en.Append(0xa8, 0x49, 0x6e, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79)
+	if err != nil {
+		return
+	}
+	if z.InMemory == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteBool(*z.InMemory)
+		if err != nil {
+			err = msgp.WrapError(err, "InMemory")
+			return
+		}
+	}
+	// write "Compress"
+	err = en.Append(0xa8, 0x43, 0x6f, 0x6d, 0x70, 0x72, 0x65, 0x73, 0x73)
+	if err != nil {
+		return
+	}
+	if z.Compress == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteBool(*z.Compress)
+		if err != nil {
+			err = msgp.WrapError(err, "Compress")
+			return
+		}
+	}
+	// write "SmallerThan"
+	err = en.Append(0xab, 0x53, 0x6d, 0x61, 0x6c, 0x6c, 0x65, 0x72, 0x54, 0x68, 0x61, 0x6e)
+	if err != nil {
+		return
+	}
+	if z.SmallerThan == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteString(*z.SmallerThan)
+		if err != nil {
+			err = msgp.WrapError(err, "SmallerThan")
+			return
+		}
+	}
+	// write "SkipErrs"
+	err = en.Append(0xa8, 0x53, 0x6b, 0x69, 0x70, 0x45, 0x72, 0x72, 0x73)
+	if err != nil {
+		return
+	}
+	if z.SkipErrs == nil {
+		err = en.WriteNil()
+		if err != nil {
+			return
+		}
+	} else {
+		err = en.WriteBool(*z.SkipErrs)
+		if err != nil {
+			err = msgp.WrapError(err, "SkipErrs")
+			return
+		}
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *BatchJobSnowball) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 6
+	// string "Disable"
+	o = append(o, 0x86, 0xa7, 0x44, 0x69, 0x73, 0x61, 0x62, 0x6c, 0x65)
+	if z.Disable == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendBool(o, *z.Disable)
+	}
+	// string "Batch"
+	o = append(o, 0xa5, 0x42, 0x61, 0x74, 0x63, 0x68)
+	if z.Batch == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendInt(o, *z.Batch)
+	}
+	// string "InMemory"
+	o = append(o, 0xa8, 0x49, 0x6e, 0x4d, 0x65, 0x6d, 0x6f, 0x72, 0x79)
+	if z.InMemory == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendBool(o, *z.InMemory)
+	}
+	// string "Compress"
+	o = append(o, 0xa8, 0x43, 0x6f, 0x6d, 0x70, 0x72, 0x65, 0x73, 0x73)
+	if z.Compress == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendBool(o, *z.Compress)
+	}
+	// string "SmallerThan"
+	o = append(o, 0xab, 0x53, 0x6d, 0x61, 0x6c, 0x6c, 0x65, 0x72, 0x54, 0x68, 0x61, 0x6e)
+	if z.SmallerThan == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendString(o, *z.SmallerThan)
+	}
+	// string "SkipErrs"
+	o = append(o, 0xa8, 0x53, 0x6b, 0x69, 0x70, 0x45, 0x72, 0x72, 0x73)
+	if z.SkipErrs == nil {
+		o = msgp.AppendNil(o)
+	} else {
+		o = msgp.AppendBool(o, *z.SkipErrs)
+	}
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *BatchJobSnowball) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "Disable":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.Disable = nil
+			} else {
+				if z.Disable == nil {
+					z.Disable = new(bool)
+				}
+				*z.Disable, bts, err = msgp.ReadBoolBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Disable")
+					return
+				}
+			}
+		case "Batch":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.Batch = nil
+			} else {
+				if z.Batch == nil {
+					z.Batch = new(int)
+				}
+				*z.Batch, bts, err = msgp.ReadIntBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Batch")
+					return
+				}
+			}
+		case "InMemory":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.InMemory = nil
+			} else {
+				if z.InMemory == nil {
+					z.InMemory = new(bool)
+				}
+				*z.InMemory, bts, err = msgp.ReadBoolBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "InMemory")
+					return
+				}
+			}
+		case "Compress":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.Compress = nil
+			} else {
+				if z.Compress == nil {
+					z.Compress = new(bool)
+				}
+				*z.Compress, bts, err = msgp.ReadBoolBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "Compress")
+					return
+				}
+			}
+		case "SmallerThan":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.SmallerThan = nil
+			} else {
+				if z.SmallerThan == nil {
+					z.SmallerThan = new(string)
+				}
+				*z.SmallerThan, bts, err = msgp.ReadStringBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "SmallerThan")
+					return
+				}
+			}
+		case "SkipErrs":
+			if msgp.IsNil(bts) {
+				bts, err = msgp.ReadNilBytes(bts)
+				if err != nil {
+					return
+				}
+				z.SkipErrs = nil
+			} else {
+				if z.SkipErrs == nil {
+					z.SkipErrs = new(bool)
+				}
+				*z.SkipErrs, bts, err = msgp.ReadBoolBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "SkipErrs")
+					return
+				}
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *BatchJobSnowball) Msgsize() (s int) {
+	s = 1 + 8
+	if z.Disable == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.BoolSize
+	}
+	s += 6
+	if z.Batch == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.IntSize
+	}
+	s += 9
+	if z.InMemory == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.BoolSize
+	}
+	s += 9
+	if z.Compress == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.BoolSize
+	}
+	s += 12
+	if z.SmallerThan == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.StringPrefixSize + len(*z.SmallerThan)
+	}
+	s += 9
+	if z.SkipErrs == nil {
+		s += msgp.NilSize
+	} else {
+		s += msgp.BoolSize
+	}
+	return
+}

cmd/batch-job-common-types_gen_test.go

@@ -347,3 +347,229 @@ func BenchmarkDecodeBatchJobRetry(b *testing.B) {
 		}
 	}
 }
+
+func TestMarshalUnmarshalBatchJobSizeFilter(t *testing.T) {
+	v := BatchJobSizeFilter{}
+	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 BenchmarkMarshalMsgBatchJobSizeFilter(b *testing.B) {
+	v := BatchJobSizeFilter{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBatchJobSizeFilter(b *testing.B) {
+	v := BatchJobSizeFilter{}
+	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 BenchmarkUnmarshalBatchJobSizeFilter(b *testing.B) {
+	v := BatchJobSizeFilter{}
+	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 TestEncodeDecodeBatchJobSizeFilter(t *testing.T) {
+	v := BatchJobSizeFilter{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBatchJobSizeFilter Msgsize() is inaccurate")
+	}
+
+	vn := BatchJobSizeFilter{}
+	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 BenchmarkEncodeBatchJobSizeFilter(b *testing.B) {
+	v := BatchJobSizeFilter{}
+	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 BenchmarkDecodeBatchJobSizeFilter(b *testing.B) {
+	v := BatchJobSizeFilter{}
+	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 TestMarshalUnmarshalBatchJobSnowball(t *testing.T) {
+	v := BatchJobSnowball{}
+	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 BenchmarkMarshalMsgBatchJobSnowball(b *testing.B) {
+	v := BatchJobSnowball{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgBatchJobSnowball(b *testing.B) {
+	v := BatchJobSnowball{}
+	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 BenchmarkUnmarshalBatchJobSnowball(b *testing.B) {
+	v := BatchJobSnowball{}
+	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 TestEncodeDecodeBatchJobSnowball(t *testing.T) {
+	v := BatchJobSnowball{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeBatchJobSnowball Msgsize() is inaccurate")
+	}
+
+	vn := BatchJobSnowball{}
+	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 BenchmarkEncodeBatchJobSnowball(b *testing.B) {
+	v := BatchJobSnowball{}
+	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 BenchmarkDecodeBatchJobSnowball(b *testing.B) {
+	v := BatchJobSnowball{}
+	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)
+		}
+	}
+}

cmd/batch-job-common-types_test.go

@@ -0,0 +1,148 @@
+// Copyright (c) 2015-2023 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 (
+	"fmt"
+	"testing"
+)
+
+func TestBatchJobSizeInRange(t *testing.T) {
+	tests := []struct {
+		objSize    int64
+		sizeFilter BatchJobSizeFilter
+		want       bool
+	}{
+		{
+			// 1Mib < 2Mib < 10MiB -> in range
+			objSize: 2 << 20,
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 10 << 20,
+				LowerBound: 1 << 20,
+			},
+			want: true,
+		},
+		{
+			// 2KiB < 1 MiB -> out of range from left
+			objSize: 2 << 10,
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 10 << 20,
+				LowerBound: 1 << 20,
+			},
+			want: false,
+		},
+		{
+			// 11MiB > 10 MiB -> out of range from right
+			objSize: 11 << 20,
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 10 << 20,
+				LowerBound: 1 << 20,
+			},
+			want: false,
+		},
+		{
+			//  2MiB < 10MiB -> in range
+			objSize: 2 << 20,
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 10 << 20,
+			},
+			want: true,
+		},
+		{
+			//  2MiB > 1MiB -> in range
+			objSize: 2 << 20,
+			sizeFilter: BatchJobSizeFilter{
+				LowerBound: 1 << 20,
+			},
+			want: true,
+		},
+	}
+
+	for i, test := range tests {
+		t.Run(fmt.Sprintf("test-%d", i+1), func(t *testing.T) {
+			if got := test.sizeFilter.InRange(test.objSize); got != test.want {
+				t.Fatalf("Expected %v but got %v", test.want, got)
+			}
+		})
+	}
+}
+
+func TestBatchJobSizeValidate(t *testing.T) {
+	errInvalidBatchJobSizeFilter := BatchJobYamlErr{
+		msg: "invalid batch-job size filter",
+	}
+
+	tests := []struct {
+		sizeFilter BatchJobSizeFilter
+		err        error
+	}{
+		{
+			// Unspecified size filter is a valid filter
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 0,
+				LowerBound: 0,
+			},
+			err: nil,
+		},
+		{
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 0,
+				LowerBound: 1 << 20,
+			},
+			err: nil,
+		},
+		{
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 10 << 20,
+				LowerBound: 0,
+			},
+			err: nil,
+		},
+		{
+			// LowerBound > UpperBound -> empty range
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 1 << 20,
+				LowerBound: 10 << 20,
+			},
+			err: errInvalidBatchJobSizeFilter,
+		},
+		{
+			// LowerBound == UpperBound -> empty range
+			sizeFilter: BatchJobSizeFilter{
+				UpperBound: 1 << 20,
+				LowerBound: 1 << 20,
+			},
+			err: errInvalidBatchJobSizeFilter,
+		},
+	}
+	for i, test := range tests {
+		t.Run(fmt.Sprintf("test-%d", i+1), func(t *testing.T) {
+			err := test.sizeFilter.Validate()
+			if err != nil {
+				gotErr := err.(BatchJobYamlErr)
+				testErr := test.err.(BatchJobYamlErr)
+				if gotErr.message() != testErr.message() {
+					t.Fatalf("Expected %v but got %v", test.err, err)
+				}
+			}
+			if err == nil && test.err != nil {
+				t.Fatalf("Expected %v but got nil", test.err)
+			}
+		})
+	}
+}

cmd/batch-replicate.go

@@ -155,6 +155,7 @@ type BatchJobReplicateSource struct {
 	Endpoint string                        `yaml:"endpoint" json:"endpoint"`
 	Path     string                        `yaml:"path" json:"path"`
 	Creds    BatchJobReplicateCredentials  `yaml:"credentials" json:"credentials"`
+	Snowball BatchJobSnowball              `yaml:"snowball" json:"snowball"`
 }
 
 // ValidPath returns true if path is valid

cmd/batch-replicate_gen.go

@@ -469,6 +469,12 @@ func (z *BatchJobReplicateSource) DecodeMsg(dc *msgp.Reader) (err error) {
 					}
 				}
 			}
+		case "Snowball":
+			err = z.Snowball.DecodeMsg(dc)
+			if err != nil {
+				err = msgp.WrapError(err, "Snowball")
+				return
+			}
 		default:
 			err = dc.Skip()
 			if err != nil {
@@ -482,9 +488,9 @@ func (z *BatchJobReplicateSource) DecodeMsg(dc *msgp.Reader) (err error) {
 
 // EncodeMsg implements msgp.Encodable
 func (z *BatchJobReplicateSource) EncodeMsg(en *msgp.Writer) (err error) {
-	// map header, size 6
+	// map header, size 7
 	// write "Type"
-	err = en.Append(0x86, 0xa4, 0x54, 0x79, 0x70, 0x65)
+	err = en.Append(0x87, 0xa4, 0x54, 0x79, 0x70, 0x65)
 	if err != nil {
 		return
 	}
@@ -569,15 +575,25 @@ func (z *BatchJobReplicateSource) EncodeMsg(en *msgp.Writer) (err error) {
 		err = msgp.WrapError(err, "Creds", "SessionToken")
 		return
 	}
+	// write "Snowball"
+	err = en.Append(0xa8, 0x53, 0x6e, 0x6f, 0x77, 0x62, 0x61, 0x6c, 0x6c)
+	if err != nil {
+		return
+	}
+	err = z.Snowball.EncodeMsg(en)
+	if err != nil {
+		err = msgp.WrapError(err, "Snowball")
+		return
+	}
 	return
 }
 
 // MarshalMsg implements msgp.Marshaler
 func (z *BatchJobReplicateSource) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.Require(b, z.Msgsize())
-	// map header, size 6
+	// map header, size 7
 	// string "Type"
-	o = append(o, 0x86, 0xa4, 0x54, 0x79, 0x70, 0x65)
+	o = append(o, 0x87, 0xa4, 0x54, 0x79, 0x70, 0x65)
 	o = msgp.AppendString(o, string(z.Type))
 	// string "Bucket"
 	o = append(o, 0xa6, 0x42, 0x75, 0x63, 0x6b, 0x65, 0x74)
@@ -603,6 +619,13 @@ func (z *BatchJobReplicateSource) MarshalMsg(b []byte) (o []byte, err error) {
 	// string "SessionToken"
 	o = append(o, 0xac, 0x53, 0x65, 0x73, 0x73, 0x69, 0x6f, 0x6e, 0x54, 0x6f, 0x6b, 0x65, 0x6e)
 	o = msgp.AppendString(o, z.Creds.SessionToken)
+	// string "Snowball"
+	o = append(o, 0xa8, 0x53, 0x6e, 0x6f, 0x77, 0x62, 0x61, 0x6c, 0x6c)
+	o, err = z.Snowball.MarshalMsg(o)
+	if err != nil {
+		err = msgp.WrapError(err, "Snowball")
+		return
+	}
 	return
 }
 
@@ -699,6 +722,12 @@ func (z *BatchJobReplicateSource) UnmarshalMsg(bts []byte) (o []byte, err error)
 					}
 				}
 			}
+		case "Snowball":
+			bts, err = z.Snowball.UnmarshalMsg(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "Snowball")
+				return
+			}
 		default:
 			bts, err = msgp.Skip(bts)
 			if err != nil {
@@ -713,7 +742,7 @@ func (z *BatchJobReplicateSource) 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 *BatchJobReplicateSource) Msgsize() (s int) {
-	s = 1 + 5 + msgp.StringPrefixSize + len(string(z.Type)) + 7 + msgp.StringPrefixSize + len(z.Bucket) + 7 + msgp.StringPrefixSize + len(z.Prefix) + 9 + msgp.StringPrefixSize + len(z.Endpoint) + 5 + msgp.StringPrefixSize + len(z.Path) + 6 + 1 + 10 + msgp.StringPrefixSize + len(z.Creds.AccessKey) + 10 + msgp.StringPrefixSize + len(z.Creds.SecretKey) + 13 + msgp.StringPrefixSize + len(z.Creds.SessionToken)
+	s = 1 + 5 + msgp.StringPrefixSize + len(string(z.Type)) + 7 + msgp.StringPrefixSize + len(z.Bucket) + 7 + msgp.StringPrefixSize + len(z.Prefix) + 9 + msgp.StringPrefixSize + len(z.Endpoint) + 5 + msgp.StringPrefixSize + len(z.Path) + 6 + 1 + 10 + msgp.StringPrefixSize + len(z.Creds.AccessKey) + 10 + msgp.StringPrefixSize + len(z.Creds.SecretKey) + 13 + msgp.StringPrefixSize + len(z.Creds.SessionToken) + 9 + z.Snowball.Msgsize()
 	return
 }
 

cmd/batch-rotate.go

@@ -154,7 +154,6 @@ type BatchJobKeyRotateV1 struct {
 	Flags      BatchJobKeyRotateFlags      `yaml:"flags" json:"flags"`
 	Bucket     string                      `yaml:"bucket" json:"bucket"`
 	Prefix     string                      `yaml:"prefix" json:"prefix"`
-	Endpoint   string                      `yaml:"endpoint" json:"endpoint"`
 	Encryption BatchJobKeyRotateEncryption `yaml:"encryption" json:"encryption"`
 }
 
@@ -261,6 +260,9 @@ func (r *BatchJobKeyRotateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 	if err := ri.load(ctx, api, job); err != nil {
 		return err
 	}
+	if ri.Complete {
+		return nil
+	}
 
 	globalBatchJobsMetrics.save(job.ID, ri)
 	lastObject := ri.Object
@@ -356,9 +358,9 @@ func (r *BatchJobKeyRotateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 	ctx, cancel := context.WithCancel(ctx)
 
 	results := make(chan ObjectInfo, 100)
-	if err := api.Walk(ctx, r.Bucket, r.Prefix, results, ObjectOptions{
-		WalkMarker: lastObject,
-		WalkFilter: skip,
+	if err := api.Walk(ctx, r.Bucket, r.Prefix, results, WalkOptions{
+		Marker: lastObject,
+		Filter: skip,
 	}); err != nil {
 		cancel()
 		// Do not need to retry if we can't list objects on source.
@@ -377,14 +379,14 @@ func (r *BatchJobKeyRotateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 			defer wk.Give()
 			for attempts := 1; attempts <= retryAttempts; attempts++ {
 				attempts := attempts
-				stopFn := globalBatchJobsMetrics.trace(batchKeyRotationMetricObject, job.ID, attempts, result)
+				stopFn := globalBatchJobsMetrics.trace(batchJobMetricKeyRotation, job.ID, attempts)
 				success := true
 				if err := r.KeyRotate(ctx, api, result); err != nil {
-					stopFn(err)
+					stopFn(result, err)
 					logger.LogIf(ctx, err)
 					success = false
 				} else {
-					stopFn(nil)
+					stopFn(result, nil)
 				}
 				ri.trackCurrentBucketObject(r.Bucket, result, success)
 				ri.RetryAttempts = attempts
@@ -398,6 +400,10 @@ func (r *BatchJobKeyRotateV1) Start(ctx context.Context, api ObjectLayer, job Ba
 					time.Sleep(delay + time.Duration(rnd.Float64()*float64(delay)))
 				}
 			}
+
+			if wait := globalBatchConfig.KeyRotationWait(); wait > 0 {
+				time.Sleep(wait)
+			}
 		}()
 	}
 	wk.Wait()

cmd/batch-rotate_gen.go

@@ -409,12 +409,6 @@ func (z *BatchJobKeyRotateV1) DecodeMsg(dc *msgp.Reader) (err error) {
 				err = msgp.WrapError(err, "Prefix")
 				return
 			}
-		case "Endpoint":
-			z.Endpoint, err = dc.ReadString()
-			if err != nil {
-				err = msgp.WrapError(err, "Endpoint")
-				return
-			}
 		case "Encryption":
 			err = z.Encryption.DecodeMsg(dc)
 			if err != nil {
@@ -434,9 +428,9 @@ func (z *BatchJobKeyRotateV1) DecodeMsg(dc *msgp.Reader) (err error) {
 
 // EncodeMsg implements msgp.Encodable
 func (z *BatchJobKeyRotateV1) EncodeMsg(en *msgp.Writer) (err error) {
-	// map header, size 6
+	// map header, size 5
 	// write "APIVersion"
-	err = en.Append(0x86, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
+	err = en.Append(0x85, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
 	if err != nil {
 		return
 	}
@@ -501,16 +495,6 @@ func (z *BatchJobKeyRotateV1) EncodeMsg(en *msgp.Writer) (err error) {
 		err = msgp.WrapError(err, "Prefix")
 		return
 	}
-	// write "Endpoint"
-	err = en.Append(0xa8, 0x45, 0x6e, 0x64, 0x70, 0x6f, 0x69, 0x6e, 0x74)
-	if err != nil {
-		return
-	}
-	err = en.WriteString(z.Endpoint)
-	if err != nil {
-		err = msgp.WrapError(err, "Endpoint")
-		return
-	}
 	// write "Encryption"
 	err = en.Append(0xaa, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x69, 0x6f, 0x6e)
 	if err != nil {
@@ -527,9 +511,9 @@ func (z *BatchJobKeyRotateV1) EncodeMsg(en *msgp.Writer) (err error) {
 // MarshalMsg implements msgp.Marshaler
 func (z *BatchJobKeyRotateV1) MarshalMsg(b []byte) (o []byte, err error) {
 	o = msgp.Require(b, z.Msgsize())
-	// map header, size 6
+	// map header, size 5
 	// string "APIVersion"
-	o = append(o, 0x86, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
+	o = append(o, 0x85, 0xaa, 0x41, 0x50, 0x49, 0x56, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e)
 	o = msgp.AppendString(o, z.APIVersion)
 	// string "Flags"
 	o = append(o, 0xa5, 0x46, 0x6c, 0x61, 0x67, 0x73)
@@ -561,9 +545,6 @@ func (z *BatchJobKeyRotateV1) MarshalMsg(b []byte) (o []byte, err error) {
 	// string "Prefix"
 	o = append(o, 0xa6, 0x50, 0x72, 0x65, 0x66, 0x69, 0x78)
 	o = msgp.AppendString(o, z.Prefix)
-	// string "Endpoint"
-	o = append(o, 0xa8, 0x45, 0x6e, 0x64, 0x70, 0x6f, 0x69, 0x6e, 0x74)
-	o = msgp.AppendString(o, z.Endpoint)
 	// string "Encryption"
 	o = append(o, 0xaa, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x69, 0x6f, 0x6e)
 	o, err = z.Encryption.MarshalMsg(o)
@@ -651,12 +632,6 @@ func (z *BatchJobKeyRotateV1) UnmarshalMsg(bts []byte) (o []byte, err error) {
 				err = msgp.WrapError(err, "Prefix")
 				return
 			}
-		case "Endpoint":
-			z.Endpoint, bts, err = msgp.ReadStringBytes(bts)
-			if err != nil {
-				err = msgp.WrapError(err, "Endpoint")
-				return
-			}
 		case "Encryption":
 			bts, err = z.Encryption.UnmarshalMsg(bts)
 			if err != nil {
@@ -677,7 +652,7 @@ func (z *BatchJobKeyRotateV1) 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 *BatchJobKeyRotateV1) Msgsize() (s int) {
-	s = 1 + 11 + msgp.StringPrefixSize + len(z.APIVersion) + 6 + 1 + 7 + z.Flags.Filter.Msgsize() + 7 + z.Flags.Notify.Msgsize() + 6 + z.Flags.Retry.Msgsize() + 7 + msgp.StringPrefixSize + len(z.Bucket) + 7 + msgp.StringPrefixSize + len(z.Prefix) + 9 + msgp.StringPrefixSize + len(z.Endpoint) + 11 + z.Encryption.Msgsize()
+	s = 1 + 11 + msgp.StringPrefixSize + len(z.APIVersion) + 6 + 1 + 7 + z.Flags.Filter.Msgsize() + 7 + z.Flags.Notify.Msgsize() + 6 + z.Flags.Retry.Msgsize() + 7 + msgp.StringPrefixSize + len(z.Bucket) + 7 + msgp.StringPrefixSize + len(z.Prefix) + 11 + z.Encryption.Msgsize()
 	return
 }
 

cmd/batchjobmetric_string.go

@@ -8,13 +8,14 @@ func _() {
 	// An "invalid array index" compiler error signifies that the constant values have changed.
 	// Re-run the stringer command to generate them again.
 	var x [1]struct{}
-	_ = x[batchReplicationMetricObject-0]
-	_ = x[batchKeyRotationMetricObject-1]
+	_ = x[batchJobMetricReplication-0]
+	_ = x[batchJobMetricKeyRotation-1]
+	_ = x[batchJobMetricExpire-2]
 }
 
-const _batchJobMetric_name = "batchReplicationMetricObjectbatchKeyRotationMetricObject"
+const _batchJobMetric_name = "ReplicationKeyRotationExpire"
 
-var _batchJobMetric_index = [...]uint8{0, 28, 56}
+var _batchJobMetric_index = [...]uint8{0, 11, 22, 28}
 
 func (i batchJobMetric) String() string {
 	if i >= batchJobMetric(len(_batchJobMetric_index)-1) {

cmd/bitrot-streaming.go

@@ -20,16 +20,12 @@ package cmd
 import (
 	"bytes"
 	"context"
-	"encoding/hex"
-	"fmt"
 	"hash"
 	"io"
-	"strings"
 	"sync"
 
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/minio/internal/ioutil"
-	"github.com/minio/minio/internal/logger"
 )
 
 // Calculates bitrot in chunks and writes the hash into the stream.
@@ -90,12 +86,12 @@ func newStreamingBitrotWriterBuffer(w io.Writer, algo BitrotAlgorithm, shardSize
 }
 
 // Returns streaming bitrot writer implementation.
-func newStreamingBitrotWriter(disk StorageAPI, volume, filePath string, length int64, algo BitrotAlgorithm, shardSize int64) io.Writer {
+func newStreamingBitrotWriter(disk StorageAPI, origvolume, volume, filePath string, length int64, algo BitrotAlgorithm, shardSize int64) io.Writer {
 	r, w := io.Pipe()
 	h := algo.New()
 
 	bw := &streamingBitrotWriter{
-		iow:          ioutil.NewDeadlineWriter(w, diskMaxTimeout),
+		iow:          ioutil.NewDeadlineWriter(w, globalDriveConfig.GetMaxTimeout()),
 		closeWithErr: w.CloseWithError,
 		h:            h,
 		shardSize:    shardSize,
@@ -110,7 +106,7 @@ func newStreamingBitrotWriter(disk StorageAPI, volume, filePath string, length i
 			bitrotSumsTotalSize := ceilFrac(length, shardSize) * int64(h.Size()) // Size used for storing bitrot checksums.
 			totalFileSize = bitrotSumsTotalSize + length
 		}
-		r.CloseWithError(disk.CreateFile(context.TODO(), volume, filePath, totalFileSize, r))
+		r.CloseWithError(disk.CreateFile(context.TODO(), origvolume, volume, filePath, totalFileSize, r))
 	}()
 	return bw
 }
@@ -134,7 +130,7 @@ func (b *streamingBitrotReader) Close() error {
 		return nil
 	}
 	if closer, ok := b.rc.(io.Closer); ok {
-		// drain the body for connection re-use at network layer.
+		// drain the body for connection reuse at network layer.
 		xhttp.DrainBody(struct {
 			io.Reader
 			io.Closer
@@ -154,29 +150,12 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 		// Can never happen unless there are programmer bugs
 		return 0, errUnexpected
 	}
-	ignoredErrs := []error{
-		errDiskNotFound,
-	}
-	if strings.HasPrefix(b.volume, minioMetaBucket) {
-		ignoredErrs = append(ignoredErrs,
-			errFileNotFound,
-			errVolumeNotFound,
-			errFileVersionNotFound,
-		)
-	}
 	if b.rc == nil {
 		// For the first ReadAt() call we need to open the stream for reading.
 		b.currOffset = offset
 		streamOffset := (offset/b.shardSize)*int64(b.h.Size()) + offset
 		if len(b.data) == 0 && b.tillOffset != streamOffset {
 			b.rc, err = b.disk.ReadFileStream(context.TODO(), b.volume, b.filePath, streamOffset, b.tillOffset-streamOffset)
-			if err != nil {
-				if !IsErr(err, ignoredErrs...) {
-					logger.LogOnceIf(GlobalContext,
-						fmt.Errorf("Reading erasure shards at (%s: %s/%s) returned '%w', will attempt to reconstruct if we have quorum",
-							b.disk, b.volume, b.filePath, err), "bitrot-read-file-stream-"+b.volume+"-"+b.filePath)
-				}
-			}
 		} else {
 			b.rc = io.NewSectionReader(bytes.NewReader(b.data), streamOffset, b.tillOffset-streamOffset)
 		}
@@ -198,10 +177,7 @@ func (b *streamingBitrotReader) ReadAt(buf []byte, offset int64) (int, error) {
 		return 0, err
 	}
 	b.h.Write(buf)
-
 	if !bytes.Equal(b.h.Sum(nil), b.hashBytes) {
-		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s  -> %s/%s - content hash does not match - expected %s, got %s",
-			b.disk, b.volume, b.filePath, hex.EncodeToString(b.hashBytes), hex.EncodeToString(b.h.Sum(nil))))
 		return 0, errFileCorrupt
 	}
 	b.currOffset += int64(len(buf))

cmd/bitrot-whole.go

@@ -19,11 +19,8 @@ package cmd
 
 import (
 	"context"
-	"fmt"
 	"hash"
 	"io"
-
-	"github.com/minio/minio/internal/logger"
 )
 
 // Implementation to calculate bitrot for the whole file.
@@ -38,12 +35,10 @@ type wholeBitrotWriter struct {
 func (b *wholeBitrotWriter) Write(p []byte) (int, error) {
 	err := b.disk.AppendFile(context.TODO(), b.volume, b.filePath, p)
 	if err != nil {
-		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s returned %w", b.disk, err))
 		return 0, err
 	}
 	_, err = b.Hash.Write(p)
 	if err != nil {
-		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s returned %w", b.disk, err))
 		return 0, err
 	}
 	return len(p), nil
@@ -72,12 +67,10 @@ func (b *wholeBitrotReader) ReadAt(buf []byte, offset int64) (n int, err error)
 	if b.buf == nil {
 		b.buf = make([]byte, b.tillOffset-offset)
 		if _, err := b.disk.ReadFile(context.TODO(), b.volume, b.filePath, offset, b.buf, b.verifier); err != nil {
-			logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, err))
 			return 0, err
 		}
 	}
 	if len(b.buf) < len(buf) {
-		logger.LogIf(GlobalContext, fmt.Errorf("Drive: %s -> %s/%s returned %w", b.disk, b.volume, b.filePath, errLessData))
 		return 0, errLessData
 	}
 	n = copy(buf, b.buf)

cmd/bitrot.go

@@ -102,9 +102,9 @@ func BitrotAlgorithmFromString(s string) (a BitrotAlgorithm) {
 	return
 }
 
-func newBitrotWriter(disk StorageAPI, volume, filePath string, length int64, algo BitrotAlgorithm, shardSize int64) io.Writer {
+func newBitrotWriter(disk StorageAPI, origvolume, volume, filePath string, length int64, algo BitrotAlgorithm, shardSize int64) io.Writer {
 	if algo == HighwayHash256S {
-		return newStreamingBitrotWriter(disk, volume, filePath, length, algo, shardSize)
+		return newStreamingBitrotWriter(disk, origvolume, volume, filePath, length, algo, shardSize)
 	}
 	return newWholeBitrotWriter(disk, volume, filePath, algo, shardSize)
 }

cmd/bitrot_test.go

@@ -36,7 +36,7 @@ func testBitrotReaderWriterAlgo(t *testing.T, bitrotAlgo BitrotAlgorithm) {
 
 	disk.MakeVol(context.Background(), volume)
 
-	writer := newBitrotWriter(disk, volume, filePath, 35, bitrotAlgo, 10)
+	writer := newBitrotWriter(disk, "", volume, filePath, 35, bitrotAlgo, 10)
 
 	_, err = writer.Write([]byte("aaaaaaaaaa"))
 	if err != nil {

cmd/bootstrap-peer-server.go

@@ -19,86 +19,81 @@ package cmd
 
 import (
 	"context"
-	"encoding/json"
+	"errors"
 	"fmt"
-	"io"
-	"net/http"
-	"net/url"
-	"os"
+	"math/rand"
 	"reflect"
+	"strings"
+	"sync"
 	"time"
 
 	"github.com/minio/minio-go/v7/pkg/set"
-	xhttp "github.com/minio/minio/internal/http"
+	"github.com/minio/minio/internal/grid"
 	"github.com/minio/minio/internal/logger"
-	"github.com/minio/minio/internal/rest"
-	"github.com/minio/mux"
 	"github.com/minio/pkg/v2/env"
 )
 
-const (
-	bootstrapRESTVersion       = "v1"
-	bootstrapRESTVersionPrefix = SlashSeparator + bootstrapRESTVersion
-	bootstrapRESTPrefix        = minioReservedBucketPath + "/bootstrap"
-	bootstrapRESTPath          = bootstrapRESTPrefix + bootstrapRESTVersionPrefix
-)
-
-const (
-	bootstrapRESTMethodHealth = "/health"
-	bootstrapRESTMethodVerify = "/verify"
-)
-
 // To abstract a node over network.
 type bootstrapRESTServer struct{}
 
+//go:generate msgp -file=$GOFILE
+
 // ServerSystemConfig - captures information about server configuration.
 type ServerSystemConfig struct {
-	MinioEndpoints EndpointServerPools
-	MinioEnv       map[string]string
+	NEndpoints int
+	CmdLines   []string
+	MinioEnv   map[string]string
 }
 
 // Diff - returns error on first difference found in two configs.
-func (s1 ServerSystemConfig) Diff(s2 ServerSystemConfig) error {
-	if s1.MinioEndpoints.NEndpoints() != s2.MinioEndpoints.NEndpoints() {
-		return fmt.Errorf("Expected number of endpoints %d, seen %d", s1.MinioEndpoints.NEndpoints(),
-			s2.MinioEndpoints.NEndpoints())
+func (s1 *ServerSystemConfig) Diff(s2 *ServerSystemConfig) error {
+	ns1 := s1.NEndpoints
+	ns2 := s2.NEndpoints
+	if ns1 != ns2 {
+		return fmt.Errorf("Expected number of endpoints %d, seen %d", ns1, ns2)
 	}
 
-	for i, ep := range s1.MinioEndpoints {
-		if ep.CmdLine != s2.MinioEndpoints[i].CmdLine {
-			return fmt.Errorf("Expected command line argument %s, seen %s", ep.CmdLine,
-				s2.MinioEndpoints[i].CmdLine)
-		}
-		if ep.SetCount != s2.MinioEndpoints[i].SetCount {
-			return fmt.Errorf("Expected set count %d, seen %d", ep.SetCount,
-				s2.MinioEndpoints[i].SetCount)
-		}
-		if ep.DrivesPerSet != s2.MinioEndpoints[i].DrivesPerSet {
-			return fmt.Errorf("Expected drives pet set %d, seen %d", ep.DrivesPerSet,
-				s2.MinioEndpoints[i].DrivesPerSet)
-		}
-		if ep.Platform != s2.MinioEndpoints[i].Platform {
-			return fmt.Errorf("Expected platform '%s', found to be on '%s'",
-				ep.Platform, s2.MinioEndpoints[i].Platform)
+	for i, cmdLine := range s1.CmdLines {
+		if cmdLine != s2.CmdLines[i] {
+			return fmt.Errorf("Expected command line argument %s, seen %s", cmdLine,
+				s2.CmdLines[i])
 		}
 	}
-	if !reflect.DeepEqual(s1.MinioEnv, s2.MinioEnv) {
-		var missing []string
-		var mismatching []string
-		for k, v := range s1.MinioEnv {
-			ev, ok := s2.MinioEnv[k]
-			if !ok {
-				missing = append(missing, k)
-			} else if v != ev {
-				mismatching = append(mismatching, k)
-			}
-		}
-		if len(mismatching) > 0 {
-			return fmt.Errorf(`Expected same MINIO_ environment variables and values across all servers: Missing environment values: %s / Mismatch environment values: %s`, missing, mismatching)
-		}
-		return fmt.Errorf(`Expected same MINIO_ environment variables and values across all servers: Missing environment values: %s`, missing)
+
+	if reflect.DeepEqual(s1.MinioEnv, s2.MinioEnv) {
+		return nil
 	}
-	return nil
+
+	// Report differences in environment variables.
+	var missing []string
+	var mismatching []string
+	for k, v := range s1.MinioEnv {
+		ev, ok := s2.MinioEnv[k]
+		if !ok {
+			missing = append(missing, k)
+		} else if v != ev {
+			mismatching = append(mismatching, k)
+		}
+	}
+	var extra []string
+	for k := range s2.MinioEnv {
+		_, ok := s1.MinioEnv[k]
+		if !ok {
+			extra = append(extra, k)
+		}
+	}
+	msg := "Expected same MINIO_ environment variables and values across all servers: "
+	if len(missing) > 0 {
+		msg += fmt.Sprintf(`Missing environment values: %v. `, missing)
+	}
+	if len(mismatching) > 0 {
+		msg += fmt.Sprintf(`Mismatching environment values: %v. `, mismatching)
+	}
+	if len(extra) > 0 {
+		msg += fmt.Sprintf(`Extra environment values: %v. `, extra)
+	}
+
+	return errors.New(strings.TrimSpace(msg))
 }
 
 var skipEnvs = map[string]struct{}{
@@ -112,7 +107,7 @@ var skipEnvs = map[string]struct{}{
 	"MINIO_SECRET_KEY":    {},
 }
 
-func getServerSystemCfg() ServerSystemConfig {
+func getServerSystemCfg() *ServerSystemConfig {
 	envs := env.List("MINIO_")
 	envValues := make(map[string]string, len(envs))
 	for _, envK := range envs {
@@ -125,131 +120,115 @@ func getServerSystemCfg() ServerSystemConfig {
 		}
 		envValues[envK] = logger.HashString(env.Get(envK, ""))
 	}
-	return ServerSystemConfig{
-		MinioEndpoints: globalEndpoints,
-		MinioEnv:       envValues,
+	scfg := &ServerSystemConfig{NEndpoints: globalEndpoints.NEndpoints(), MinioEnv: envValues}
+	var cmdLines []string
+	for _, ep := range globalEndpoints {
+		cmdLines = append(cmdLines, ep.CmdLine)
 	}
+	scfg.CmdLines = cmdLines
+	return scfg
 }
 
-func (b *bootstrapRESTServer) writeErrorResponse(w http.ResponseWriter, err error) {
-	w.WriteHeader(http.StatusForbidden)
-	w.Write([]byte(err.Error()))
+func (s *bootstrapRESTServer) VerifyHandler(params *grid.MSS) (*ServerSystemConfig, *grid.RemoteErr) {
+	return getServerSystemCfg(), nil
 }
 
-// HealthHandler returns success if request is valid
-func (b *bootstrapRESTServer) HealthHandler(w http.ResponseWriter, r *http.Request) {}
-
-func (b *bootstrapRESTServer) VerifyHandler(w http.ResponseWriter, r *http.Request) {
-	ctx := newContext(r, w, "VerifyHandler")
-
-	if err := storageServerRequestValidate(r); err != nil {
-		b.writeErrorResponse(w, err)
-		return
-	}
-
-	cfg := getServerSystemCfg()
-	logger.LogIf(ctx, json.NewEncoder(w).Encode(&cfg))
-}
+var serverVerifyHandler = grid.NewSingleHandler[*grid.MSS, *ServerSystemConfig](grid.HandlerServerVerify, grid.NewMSS, func() *ServerSystemConfig { return &ServerSystemConfig{} })
 
 // registerBootstrapRESTHandlers - register bootstrap rest router.
-func registerBootstrapRESTHandlers(router *mux.Router) {
-	h := func(f http.HandlerFunc) http.HandlerFunc {
-		return collectInternodeStats(httpTraceHdrs(f))
-	}
-
+func registerBootstrapRESTHandlers(gm *grid.Manager) {
 	server := &bootstrapRESTServer{}
-	subrouter := router.PathPrefix(bootstrapRESTPrefix).Subrouter()
-
-	subrouter.Methods(http.MethodPost).Path(bootstrapRESTVersionPrefix + bootstrapRESTMethodHealth).HandlerFunc(
-		h(server.HealthHandler))
-
-	subrouter.Methods(http.MethodPost).Path(bootstrapRESTVersionPrefix + bootstrapRESTMethodVerify).HandlerFunc(
-		h(server.VerifyHandler))
+	logger.FatalIf(serverVerifyHandler.Register(gm, server.VerifyHandler), "unable to register handler")
 }
 
 // client to talk to bootstrap NEndpoints.
 type bootstrapRESTClient struct {
-	endpoint   Endpoint
-	restClient *rest.Client
+	gridConn *grid.Connection
 }
 
-// Wrapper to restClient.Call to handle network errors, in case of network error the connection is marked disconnected
-// permanently. The only way to restore the connection is at the xl-sets layer by xlsets.monitorAndConnectEndpoints()
-// after verifying format.json
-func (client *bootstrapRESTClient) callWithContext(ctx context.Context, method string, values url.Values, body io.Reader, length int64) (respBody io.ReadCloser, err error) {
-	if values == nil {
-		values = make(url.Values)
+// Verify function verifies the server config.
+func (client *bootstrapRESTClient) Verify(ctx context.Context, srcCfg *ServerSystemConfig) (err error) {
+	if newObjectLayerFn() != nil {
+		return nil
 	}
 
-	respBody, err = client.restClient.Call(ctx, method, values, body, length)
-	if err == nil {
-		return respBody, nil
+	recvCfg, err := serverVerifyHandler.Call(ctx, client.gridConn, grid.NewMSS())
+	if err != nil {
+		return err
 	}
+	// We do not need the response after returning.
+	defer serverVerifyHandler.PutResponse(recvCfg)
 
-	return nil, err
+	return srcCfg.Diff(recvCfg)
 }
 
 // Stringer provides a canonicalized representation of node.
 func (client *bootstrapRESTClient) String() string {
-	return client.endpoint.String()
+	return client.gridConn.String()
 }
 
-// Verify - fetches system server config.
-func (client *bootstrapRESTClient) Verify(ctx context.Context, srcCfg ServerSystemConfig) (err error) {
-	if newObjectLayerFn() != nil {
-		return nil
-	}
-	respBody, err := client.callWithContext(ctx, bootstrapRESTMethodVerify, nil, nil, -1)
-	if err != nil {
-		return
-	}
-	defer xhttp.DrainBody(respBody)
-	recvCfg := ServerSystemConfig{}
-	if err = json.NewDecoder(respBody).Decode(&recvCfg); err != nil {
-		return err
-	}
-	return srcCfg.Diff(recvCfg)
-}
-
-func verifyServerSystemConfig(ctx context.Context, endpointServerPools EndpointServerPools) error {
+func verifyServerSystemConfig(ctx context.Context, endpointServerPools EndpointServerPools, gm *grid.Manager) error {
 	srcCfg := getServerSystemCfg()
-	clnts := newBootstrapRESTClients(endpointServerPools)
+	clnts := newBootstrapRESTClients(endpointServerPools, gm)
 	var onlineServers int
 	var offlineEndpoints []error
 	var incorrectConfigs []error
 	var retries int
+	var mu sync.Mutex
 	for onlineServers < len(clnts)/2 {
+		var wg sync.WaitGroup
+		wg.Add(len(clnts))
+		onlineServers = 0
 		for _, clnt := range clnts {
-			if err := clnt.Verify(ctx, srcCfg); err != nil {
-				bootstrapTraceMsg(fmt.Sprintf("clnt.Verify: %v, endpoint: %v", err, clnt.endpoint))
-				if !isNetworkError(err) {
-					logger.LogOnceIf(ctx, fmt.Errorf("%s has incorrect configuration: %w", clnt.String(), err), clnt.String())
-					incorrectConfigs = append(incorrectConfigs, fmt.Errorf("%s has incorrect configuration: %w", clnt.String(), err))
-				} else {
-					offlineEndpoints = append(offlineEndpoints, fmt.Errorf("%s is unreachable: %w", clnt.String(), err))
+			go func(clnt *bootstrapRESTClient) {
+				defer wg.Done()
+
+				if clnt.gridConn.State() != grid.StateConnected {
+					mu.Lock()
+					offlineEndpoints = append(offlineEndpoints, fmt.Errorf("%s is unreachable: %w", clnt, grid.ErrDisconnected))
+					mu.Unlock()
+					return
 				}
-				continue
-			}
-			onlineServers++
+
+				ctx, cancel := context.WithTimeout(ctx, 2*time.Second)
+				defer cancel()
+
+				err := clnt.Verify(ctx, srcCfg)
+				mu.Lock()
+				if err != nil {
+					bootstrapTraceMsg(fmt.Sprintf("clnt.Verify: %v, endpoint: %s", err, clnt))
+					if !isNetworkError(err) {
+						logger.LogOnceIf(context.Background(), fmt.Errorf("%s has incorrect configuration: %w", clnt, err), "incorrect_"+clnt.String())
+						incorrectConfigs = append(incorrectConfigs, fmt.Errorf("%s has incorrect configuration: %w", clnt, err))
+					} else {
+						offlineEndpoints = append(offlineEndpoints, fmt.Errorf("%s is unreachable: %w", clnt, err))
+					}
+				} else {
+					onlineServers++
+				}
+				mu.Unlock()
+			}(clnt)
 		}
+		wg.Wait()
+
 		select {
 		case <-ctx.Done():
 			return ctx.Err()
 		default:
-			// Sleep for a while - so that we don't go into
-			// 100% CPU when half the endpoints are offline.
-			time.Sleep(100 * time.Millisecond)
+			// Sleep and stagger to avoid blocked CPU and thundering
+			// herd upon start up sequence.
+			time.Sleep(25*time.Millisecond + time.Duration(rand.Int63n(int64(100*time.Millisecond))))
 			retries++
 			// after 20 retries start logging that servers are not reachable yet
 			if retries >= 20 {
-				logger.Info(fmt.Sprintf("Waiting for atleast %d remote servers with valid configuration to be online", len(clnts)/2))
+				logger.Info(fmt.Sprintf("Waiting for at least %d remote servers with valid configuration to be online", len(clnts)/2))
 				if len(offlineEndpoints) > 0 {
 					logger.Info(fmt.Sprintf("Following servers are currently offline or unreachable %s", offlineEndpoints))
 				}
 				if len(incorrectConfigs) > 0 {
 					logger.Info(fmt.Sprintf("Following servers have mismatching configuration %s", incorrectConfigs))
 				}
-				retries = 0 // reset to log again after 5 retries.
+				retries = 0 // reset to log again after 20 retries.
 			}
 			offlineEndpoints = nil
 			incorrectConfigs = nil
@@ -258,39 +237,20 @@ func verifyServerSystemConfig(ctx context.Context, endpointServerPools EndpointS
 	return nil
 }
 
-func newBootstrapRESTClients(endpointServerPools EndpointServerPools) []*bootstrapRESTClient {
-	seenHosts := set.NewStringSet()
+func newBootstrapRESTClients(endpointServerPools EndpointServerPools, gm *grid.Manager) []*bootstrapRESTClient {
+	seenClient := set.NewStringSet()
 	var clnts []*bootstrapRESTClient
 	for _, ep := range endpointServerPools {
 		for _, endpoint := range ep.Endpoints {
-			if seenHosts.Contains(endpoint.Host) {
+			if endpoint.IsLocal {
 				continue
 			}
-			seenHosts.Add(endpoint.Host)
-
-			// Only proceed for remote endpoints.
-			if !endpoint.IsLocal {
-				cl := newBootstrapRESTClient(endpoint)
-				if serverDebugLog {
-					cl.restClient.TraceOutput = os.Stdout
-				}
-				clnts = append(clnts, cl)
+			if seenClient.Contains(endpoint.Host) {
+				continue
 			}
+			seenClient.Add(endpoint.Host)
+			clnts = append(clnts, &bootstrapRESTClient{gm.Connection(endpoint.GridHost())})
 		}
 	}
 	return clnts
 }
-
-// Returns a new bootstrap client.
-func newBootstrapRESTClient(endpoint Endpoint) *bootstrapRESTClient {
-	serverURL := &url.URL{
-		Scheme: endpoint.Scheme,
-		Host:   endpoint.Host,
-		Path:   bootstrapRESTPath,
-	}
-
-	restClient := rest.NewClient(serverURL, globalInternodeTransport, newCachedAuthToken())
-	restClient.HealthCheckFn = nil
-
-	return &bootstrapRESTClient{endpoint: endpoint, restClient: restClient}
-}

cmd/bootstrap-peer-server_gen.go

@@ -0,0 +1,270 @@
+package cmd
+
+// Code generated by github.com/tinylib/msgp DO NOT EDIT.
+
+import (
+	"github.com/tinylib/msgp/msgp"
+)
+
+// DecodeMsg implements msgp.Decodable
+func (z *ServerSystemConfig) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "NEndpoints":
+			z.NEndpoints, err = dc.ReadInt()
+			if err != nil {
+				err = msgp.WrapError(err, "NEndpoints")
+				return
+			}
+		case "CmdLines":
+			var zb0002 uint32
+			zb0002, err = dc.ReadArrayHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "CmdLines")
+				return
+			}
+			if cap(z.CmdLines) >= int(zb0002) {
+				z.CmdLines = (z.CmdLines)[:zb0002]
+			} else {
+				z.CmdLines = make([]string, zb0002)
+			}
+			for za0001 := range z.CmdLines {
+				z.CmdLines[za0001], err = dc.ReadString()
+				if err != nil {
+					err = msgp.WrapError(err, "CmdLines", za0001)
+					return
+				}
+			}
+		case "MinioEnv":
+			var zb0003 uint32
+			zb0003, err = dc.ReadMapHeader()
+			if err != nil {
+				err = msgp.WrapError(err, "MinioEnv")
+				return
+			}
+			if z.MinioEnv == nil {
+				z.MinioEnv = make(map[string]string, zb0003)
+			} else if len(z.MinioEnv) > 0 {
+				for key := range z.MinioEnv {
+					delete(z.MinioEnv, key)
+				}
+			}
+			for zb0003 > 0 {
+				zb0003--
+				var za0002 string
+				var za0003 string
+				za0002, err = dc.ReadString()
+				if err != nil {
+					err = msgp.WrapError(err, "MinioEnv")
+					return
+				}
+				za0003, err = dc.ReadString()
+				if err != nil {
+					err = msgp.WrapError(err, "MinioEnv", za0002)
+					return
+				}
+				z.MinioEnv[za0002] = za0003
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *ServerSystemConfig) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 3
+	// write "NEndpoints"
+	err = en.Append(0x83, 0xaa, 0x4e, 0x45, 0x6e, 0x64, 0x70, 0x6f, 0x69, 0x6e, 0x74, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteInt(z.NEndpoints)
+	if err != nil {
+		err = msgp.WrapError(err, "NEndpoints")
+		return
+	}
+	// write "CmdLines"
+	err = en.Append(0xa8, 0x43, 0x6d, 0x64, 0x4c, 0x69, 0x6e, 0x65, 0x73)
+	if err != nil {
+		return
+	}
+	err = en.WriteArrayHeader(uint32(len(z.CmdLines)))
+	if err != nil {
+		err = msgp.WrapError(err, "CmdLines")
+		return
+	}
+	for za0001 := range z.CmdLines {
+		err = en.WriteString(z.CmdLines[za0001])
+		if err != nil {
+			err = msgp.WrapError(err, "CmdLines", za0001)
+			return
+		}
+	}
+	// write "MinioEnv"
+	err = en.Append(0xa8, 0x4d, 0x69, 0x6e, 0x69, 0x6f, 0x45, 0x6e, 0x76)
+	if err != nil {
+		return
+	}
+	err = en.WriteMapHeader(uint32(len(z.MinioEnv)))
+	if err != nil {
+		err = msgp.WrapError(err, "MinioEnv")
+		return
+	}
+	for za0002, za0003 := range z.MinioEnv {
+		err = en.WriteString(za0002)
+		if err != nil {
+			err = msgp.WrapError(err, "MinioEnv")
+			return
+		}
+		err = en.WriteString(za0003)
+		if err != nil {
+			err = msgp.WrapError(err, "MinioEnv", za0002)
+			return
+		}
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *ServerSystemConfig) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 3
+	// string "NEndpoints"
+	o = append(o, 0x83, 0xaa, 0x4e, 0x45, 0x6e, 0x64, 0x70, 0x6f, 0x69, 0x6e, 0x74, 0x73)
+	o = msgp.AppendInt(o, z.NEndpoints)
+	// string "CmdLines"
+	o = append(o, 0xa8, 0x43, 0x6d, 0x64, 0x4c, 0x69, 0x6e, 0x65, 0x73)
+	o = msgp.AppendArrayHeader(o, uint32(len(z.CmdLines)))
+	for za0001 := range z.CmdLines {
+		o = msgp.AppendString(o, z.CmdLines[za0001])
+	}
+	// string "MinioEnv"
+	o = append(o, 0xa8, 0x4d, 0x69, 0x6e, 0x69, 0x6f, 0x45, 0x6e, 0x76)
+	o = msgp.AppendMapHeader(o, uint32(len(z.MinioEnv)))
+	for za0002, za0003 := range z.MinioEnv {
+		o = msgp.AppendString(o, za0002)
+		o = msgp.AppendString(o, za0003)
+	}
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *ServerSystemConfig) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "NEndpoints":
+			z.NEndpoints, bts, err = msgp.ReadIntBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "NEndpoints")
+				return
+			}
+		case "CmdLines":
+			var zb0002 uint32
+			zb0002, bts, err = msgp.ReadArrayHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "CmdLines")
+				return
+			}
+			if cap(z.CmdLines) >= int(zb0002) {
+				z.CmdLines = (z.CmdLines)[:zb0002]
+			} else {
+				z.CmdLines = make([]string, zb0002)
+			}
+			for za0001 := range z.CmdLines {
+				z.CmdLines[za0001], bts, err = msgp.ReadStringBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "CmdLines", za0001)
+					return
+				}
+			}
+		case "MinioEnv":
+			var zb0003 uint32
+			zb0003, bts, err = msgp.ReadMapHeaderBytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "MinioEnv")
+				return
+			}
+			if z.MinioEnv == nil {
+				z.MinioEnv = make(map[string]string, zb0003)
+			} else if len(z.MinioEnv) > 0 {
+				for key := range z.MinioEnv {
+					delete(z.MinioEnv, key)
+				}
+			}
+			for zb0003 > 0 {
+				var za0002 string
+				var za0003 string
+				zb0003--
+				za0002, bts, err = msgp.ReadStringBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "MinioEnv")
+					return
+				}
+				za0003, bts, err = msgp.ReadStringBytes(bts)
+				if err != nil {
+					err = msgp.WrapError(err, "MinioEnv", za0002)
+					return
+				}
+				z.MinioEnv[za0002] = za0003
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *ServerSystemConfig) Msgsize() (s int) {
+	s = 1 + 11 + msgp.IntSize + 9 + msgp.ArrayHeaderSize
+	for za0001 := range z.CmdLines {
+		s += msgp.StringPrefixSize + len(z.CmdLines[za0001])
+	}
+	s += 9 + msgp.MapHeaderSize
+	if z.MinioEnv != nil {
+		for za0002, za0003 := range z.MinioEnv {
+			_ = za0003
+			s += msgp.StringPrefixSize + len(za0002) + msgp.StringPrefixSize + len(za0003)
+		}
+	}
+	return
+}

cmd/bootstrap-peer-server_gen_test.go

@@ -0,0 +1,123 @@
+package cmd
+
+// Code generated by github.com/tinylib/msgp DO NOT EDIT.
+
+import (
+	"bytes"
+	"testing"
+
+	"github.com/tinylib/msgp/msgp"
+)
+
+func TestMarshalUnmarshalServerSystemConfig(t *testing.T) {
+	v := ServerSystemConfig{}
+	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 BenchmarkMarshalMsgServerSystemConfig(b *testing.B) {
+	v := ServerSystemConfig{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgServerSystemConfig(b *testing.B) {
+	v := ServerSystemConfig{}
+	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 BenchmarkUnmarshalServerSystemConfig(b *testing.B) {
+	v := ServerSystemConfig{}
+	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 TestEncodeDecodeServerSystemConfig(t *testing.T) {
+	v := ServerSystemConfig{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeServerSystemConfig Msgsize() is inaccurate")
+	}
+
+	vn := ServerSystemConfig{}
+	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 BenchmarkEncodeServerSystemConfig(b *testing.B) {
+	v := ServerSystemConfig{}
+	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 BenchmarkDecodeServerSystemConfig(b *testing.B) {
+	v := ServerSystemConfig{}
+	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)
+		}
+	}
+}

cmd/bucket-handlers.go

@@ -51,6 +51,7 @@ import (
 	sse "github.com/minio/minio/internal/bucket/encryption"
 	objectlock "github.com/minio/minio/internal/bucket/object/lock"
 	"github.com/minio/minio/internal/bucket/replication"
+	"github.com/minio/minio/internal/config/cache"
 	"github.com/minio/minio/internal/config/dns"
 	"github.com/minio/minio/internal/crypto"
 	"github.com/minio/minio/internal/event"
@@ -423,7 +424,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		return
 	}
 
-	// Content-Md5 is requied should be set
+	// Content-Md5 is required should be set
 	// http://docs.aws.amazon.com/AmazonS3/latest/API/multiobjectdeleteapi.html
 	if _, ok := r.Header[xhttp.ContentMD5]; !ok {
 		writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrMissingContentMD5), r.URL)
@@ -466,17 +467,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 	// Ignore errors here to preserve the S3 error behavior of GetBucketInfo()
 	checkRequestAuthType(ctx, r, policy.DeleteObjectAction, bucket, "")
 
-	// Before proceeding validate if bucket exists.
-	_, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{})
-	if err != nil {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
-	}
-
 	deleteObjectsFn := objectAPI.DeleteObjects
-	if api.CacheAPI() != nil {
-		deleteObjectsFn = api.CacheAPI().DeleteObjects
-	}
 
 	// Return Malformed XML as S3 spec if the number of objects is empty
 	if len(deleteObjectsReq.Objects) == 0 || len(deleteObjectsReq.Objects) > maxDeleteList {
@@ -486,9 +477,6 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 
 	objectsToDelete := map[ObjectToDelete]int{}
 	getObjectInfoFn := objectAPI.GetObjectInfo
-	if api.CacheAPI() != nil {
-		getObjectInfoFn = api.CacheAPI().GetObjectInfo
-	}
 
 	var (
 		hasLockEnabled bool
@@ -616,6 +604,12 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		VersionSuspended: vc.Suspended(),
 	})
 
+	// Are all objects saying bucket not found?
+	if isAllBucketsNotFound(errs) {
+		writeErrorResponse(ctx, w, toAPIError(ctx, errs[0]), r.URL)
+		return
+	}
+
 	for i := range errs {
 		// DeleteMarkerVersionID is not used specifically to avoid
 		// lookup errors, since DeleteMarkerVersionID is only
@@ -623,7 +617,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		// specify a versionID.
 		objToDel := ObjectToDelete{
 			ObjectV: ObjectV{
-				ObjectName: dObjects[i].ObjectName,
+				ObjectName: decodeDirObject(dObjects[i].ObjectName),
 				VersionID:  dObjects[i].VersionID,
 			},
 			VersionPurgeStatus:            dObjects[i].VersionPurgeStatus(),
@@ -674,6 +668,8 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 			continue
 		}
 
+		defer globalCacheConfig.Delete(bucket, dobj.ObjectName)
+
 		if replicateDeletes && (dobj.DeleteMarkerReplicationStatus() == replication.Pending || dobj.VersionPurgeStatus() == Pending) {
 			// copy so we can re-add null ID.
 			dobj := dobj
@@ -716,7 +712,7 @@ func (api objectAPIHandlers) DeleteMultipleObjectsHandler(w http.ResponseWriter,
 		if os == nil { // skip objects that weren't deleted due to invalid versionID etc.
 			continue
 		}
-		logger.LogIf(ctx, os.Sweep())
+		os.Sweep()
 	}
 }
 
@@ -794,7 +790,7 @@ func (api objectAPIHandlers) PutBucketHandler(w http.ResponseWriter, r *http.Req
 
 	// check if client is attempting to create more buckets, complain about it.
 	if currBuckets := globalBucketMetadataSys.Count(); currBuckets+1 > maxBuckets {
-		logger.LogIf(ctx, fmt.Errorf("An attempt to create %d buckets beyond recommended %d", currBuckets+1, maxBuckets))
+		logger.LogIf(ctx, fmt.Errorf("Please avoid creating more buckets %d beyond recommended %d", currBuckets+1, maxBuckets))
 	}
 
 	opts := MakeBucketOptions{
@@ -989,7 +985,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		}
 
 		var b bytes.Buffer
-		if fileName == "" {
+		if name != "file" {
 			if http.CanonicalHeaderKey(name) == http.CanonicalHeaderKey("x-minio-fanout-list") {
 				dec := json.NewDecoder(part)
 
@@ -1045,11 +1041,14 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		break
 	}
 
-	if _, ok := formValues["Key"]; !ok {
+	if keyName, ok := formValues["Key"]; !ok {
 		apiErr := errorCodes.ToAPIErr(ErrMalformedPOSTRequest)
 		apiErr.Description = fmt.Sprintf("%s (%v)", apiErr.Description, errors.New("The name of the uploaded key is missing"))
 		writeErrorResponse(ctx, w, apiErr, r.URL)
 		return
+	} else if fileName == "" && len(keyName) >= 1 {
+		// if we can't get fileName. We use keyName[0] to fileName
+		fileName = keyName[0]
 	}
 
 	if fileName == "" {
@@ -1141,7 +1140,6 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 
 	hashReader, err := hash.NewReader(ctx, reader, fileSize, "", "", fileSize)
 	if err != nil {
-		logger.LogIf(ctx, err)
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
@@ -1196,7 +1194,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 	})
 
 	var opts ObjectOptions
-	opts, err = putOpts(ctx, r, bucket, object, metadata)
+	opts, err = putOptsFromReq(ctx, r, bucket, object, metadata)
 	if err != nil {
 		writeErrorResponseHeadersOnly(w, toAPIError(ctx, err))
 		return
@@ -1347,6 +1345,22 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 					continue
 				}
 
+				asize, err := objInfo.GetActualSize()
+				if err != nil {
+					asize = objInfo.Size
+				}
+
+				globalCacheConfig.Set(&cache.ObjectInfo{
+					Key:          objInfo.Name,
+					Bucket:       objInfo.Bucket,
+					ETag:         getDecryptedETag(formValues, objInfo, false),
+					ModTime:      objInfo.ModTime,
+					Expires:      objInfo.Expires.UTC().Format(http.TimeFormat),
+					CacheControl: objInfo.CacheControl,
+					Metadata:     cleanReservedKeys(objInfo.UserDefined),
+					Size:         asize,
+				})
+
 				fanOutResp = append(fanOutResp, minio.PutObjectFanOutResponse{
 					Key:          objInfo.Name,
 					ETag:         getDecryptedETag(formValues, objInfo, false),
@@ -1380,7 +1394,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 			// Notify object created events.
 			sendEvent(eventArgsList[i])
 
-			if eventArgsList[i].Object.NumVersions > dataScannerExcessiveVersionsThreshold {
+			if eventArgsList[i].Object.NumVersions > int(scannerExcessObjectVersions.Load()) {
 				// Send events for excessive versions.
 				sendEvent(eventArgs{
 					EventName:    event.ObjectManyVersions,
@@ -1391,6 +1405,15 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 					UserAgent:    r.UserAgent() + " " + "MinIO-Fan-Out",
 					Host:         handlers.GetSourceIP(r),
 				})
+
+				auditLogInternal(context.Background(), AuditLogOptions{
+					Event:     "scanner:manyversions",
+					APIName:   "PostPolicyBucket",
+					Bucket:    eventArgsList[i].Object.Bucket,
+					Object:    eventArgsList[i].Object.Name,
+					VersionID: eventArgsList[i].Object.VersionID,
+					Status:    http.StatusText(http.StatusOK),
+				})
 			}
 		}
 
@@ -1403,10 +1426,12 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		return
 	}
 
+	etag := getDecryptedETag(formValues, objInfo, false)
+
 	// We must not use the http.Header().Set method here because some (broken)
 	// clients expect the ETag header key to be literally "ETag" - not "Etag" (case-sensitive).
 	// Therefore, we have to set the ETag directly as map entry.
-	w.Header()[xhttp.ETag] = []string{`"` + objInfo.ETag + `"`}
+	w.Header()[xhttp.ETag] = []string{`"` + etag + `"`}
 
 	// Set the relevant version ID as part of the response header.
 	if objInfo.VersionID != "" && objInfo.VersionID != nullVersionID {
@@ -1417,6 +1442,22 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		w.Header().Set(xhttp.Location, obj)
 	}
 
+	asize, err := objInfo.GetActualSize()
+	if err != nil {
+		asize = objInfo.Size
+	}
+
+	defer globalCacheConfig.Set(&cache.ObjectInfo{
+		Key:          objInfo.Name,
+		Bucket:       objInfo.Bucket,
+		ETag:         etag,
+		ModTime:      objInfo.ModTime,
+		Expires:      objInfo.ExpiresStr(),
+		CacheControl: objInfo.CacheControl,
+		Metadata:     cleanReservedKeys(objInfo.UserDefined),
+		Size:         asize,
+	})
+
 	// Notify object created event.
 	defer sendEvent(eventArgs{
 		EventName:    event.ObjectCreatedPost,
@@ -1428,7 +1469,7 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 		Host:         handlers.GetSourceIP(r),
 	})
 
-	if objInfo.NumVersions > dataScannerExcessiveVersionsThreshold {
+	if objInfo.NumVersions > int(scannerExcessObjectVersions.Load()) {
 		defer sendEvent(eventArgs{
 			EventName:    event.ObjectManyVersions,
 			BucketName:   objInfo.Bucket,
@@ -1438,6 +1479,15 @@ func (api objectAPIHandlers) PostPolicyBucketHandler(w http.ResponseWriter, r *h
 			UserAgent:    r.UserAgent(),
 			Host:         handlers.GetSourceIP(r),
 		})
+
+		auditLogInternal(context.Background(), AuditLogOptions{
+			Event:     "scanner:manyversions",
+			APIName:   "PostPolicyBucket",
+			Bucket:    objInfo.Bucket,
+			Object:    objInfo.Name,
+			VersionID: objInfo.VersionID,
+			Status:    http.StatusText(http.StatusOK),
+		})
 	}
 
 	if redirectURL != nil { // success_action_redirect is valid and set.
@@ -1624,9 +1674,11 @@ func (api objectAPIHandlers) DeleteBucketHandler(w http.ResponseWriter, r *http.
 	}
 
 	// Return an error if the bucket does not exist
-	if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil && !forceDelete {
-		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
-		return
+	if !forceDelete {
+		if _, err := objectAPI.GetBucketInfo(ctx, bucket, BucketOptions{}); err != nil {
+			writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+			return
+		}
 	}
 
 	// Attempt to delete bucket.

cmd/bucket-handlers_test.go

@@ -252,7 +252,7 @@ func testHeadBucketHandler(obj ObjectLayer, instanceType, bucketName string, api
 		},
 		// Test case - 3.
 		// Testing for signature mismatch error.
-		// setting invalid acess and secret key.
+		// setting invalid access and secret key.
 		{
 			bucketName:         bucketName,
 			accessKey:          "abcd",
@@ -415,7 +415,7 @@ func testListMultipartUploadsHandler(obj ObjectLayer, instanceType, bucketName s
 			shouldPass:         false,
 		},
 		// Test case - 6.
-		// Setting a negative value to max-uploads paramater, should result in http.StatusBadRequest.
+		// Setting a negative value to max-uploads parameter, should result in http.StatusBadRequest.
 		{
 			bucket:             bucketName,
 			prefix:             "",
@@ -880,6 +880,15 @@ func testAPIDeleteMultipleObjectsHandler(obj ObjectLayer, instanceType, bucketNa
 			expectedContent:    encodedAnonResponseWithPartialPublicAccess,
 			expectedRespStatus: http.StatusOK,
 		},
+		// Test case - 7.
+		// Bucket does not exist.
+		7: {
+			bucket:             "unknown-bucket-name",
+			objects:            successRequest0,
+			accessKey:          credentials.AccessKey,
+			secretKey:          credentials.SecretKey,
+			expectedRespStatus: http.StatusNotFound,
+		},
 	}
 
 	for i, testCase := range testCases {
@@ -888,13 +897,12 @@ func testAPIDeleteMultipleObjectsHandler(obj ObjectLayer, instanceType, bucketNa
 
 		// Generate a signed or anonymous request based on the testCase
 		if testCase.accessKey != "" {
-			req, err = newTestSignedRequestV4(http.MethodPost, getDeleteMultipleObjectsURL("", bucketName),
+			req, err = newTestSignedRequestV4(http.MethodPost, getDeleteMultipleObjectsURL("", testCase.bucket),
 				int64(len(testCase.objects)), bytes.NewReader(testCase.objects), testCase.accessKey, testCase.secretKey, nil)
 		} else {
-			req, err = newTestRequest(http.MethodPost, getDeleteMultipleObjectsURL("", bucketName),
+			req, err = newTestRequest(http.MethodPost, getDeleteMultipleObjectsURL("", testCase.bucket),
 				int64(len(testCase.objects)), bytes.NewReader(testCase.objects))
 		}
-
 		if err != nil {
 			t.Fatalf("Failed to create HTTP request for DeleteMultipleObjects: <ERROR> %v", err)
 		}

cmd/bucket-lifecycle-audit.go

@@ -25,6 +25,7 @@ type lcEventSrc uint8
 //revive:disable:var-naming Underscores is used here to indicate where common prefix ends and the enumeration name begins
 const (
 	lcEventSrc_None lcEventSrc = iota
+	lcEventSrc_Heal
 	lcEventSrc_Scanner
 	lcEventSrc_Decom
 	lcEventSrc_Rebal

cmd/bucket-lifecycle-handlers.go

@@ -22,6 +22,7 @@ import (
 	"io"
 	"net/http"
 	"strconv"
+	"time"
 
 	"github.com/minio/minio/internal/bucket/lifecycle"
 	xhttp "github.com/minio/minio/internal/http"
@@ -86,6 +87,41 @@ func (api objectAPIHandlers) PutBucketLifecycleHandler(w http.ResponseWriter, r
 		return
 	}
 
+	// Create a map of updated set of rules in request
+	updatedRules := make(map[string]lifecycle.Rule, len(bucketLifecycle.Rules))
+	for _, rule := range bucketLifecycle.Rules {
+		updatedRules[rule.ID] = rule
+	}
+
+	// Get list of rules for the bucket from disk
+	meta, err := globalBucketMetadataSys.GetConfigFromDisk(ctx, bucket)
+	if err != nil {
+		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+		return
+	}
+	expiryRuleRemoved := false
+	if len(meta.LifecycleConfigXML) > 0 {
+		var lcCfg lifecycle.Lifecycle
+		if err := xml.Unmarshal(meta.LifecycleConfigXML, &lcCfg); err != nil {
+			writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
+			return
+		}
+		for _, rl := range lcCfg.Rules {
+			updRule, ok := updatedRules[rl.ID]
+			// original rule had expiry that is no longer in the new config,
+			// or rule is present but missing expiration flags
+			if (!rl.Expiration.IsNull() || !rl.NoncurrentVersionExpiration.IsNull()) &&
+				(!ok || (updRule.Expiration.IsNull() && updRule.NoncurrentVersionExpiration.IsNull())) {
+				expiryRuleRemoved = true
+			}
+		}
+	}
+
+	if bucketLifecycle.HasExpiry() || expiryRuleRemoved {
+		currtime := time.Now()
+		bucketLifecycle.ExpiryUpdatedAt = &currtime
+	}
+
 	configData, err := xml.Marshal(bucketLifecycle)
 	if err != nil {
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
@@ -142,6 +178,8 @@ func (api objectAPIHandlers) GetBucketLifecycleHandler(w http.ResponseWriter, r
 		writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL)
 		return
 	}
+	// explicitly set ExpiryUpdatedAt nil as its meant for internal consumption only
+	config.ExpiryUpdatedAt = nil
 
 	configData, err := xml.Marshal(config)
 	if err != nil {

cmd/bucket-lifecycle.go

@@ -39,10 +39,13 @@ import (
 	"github.com/minio/minio/internal/bucket/lifecycle"
 	"github.com/minio/minio/internal/event"
 	xhttp "github.com/minio/minio/internal/http"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/logger"
 	"github.com/minio/minio/internal/s3select"
 	"github.com/minio/pkg/v2/env"
+	xnet "github.com/minio/pkg/v2/net"
 	"github.com/minio/pkg/v2/workers"
+	"github.com/zeebo/xxh3"
 )
 
 const (
@@ -116,8 +119,8 @@ func (es *expiryState) PendingTasks() int {
 // close closes work channels exactly once.
 func (es *expiryState) close() {
 	es.once.Do(func() {
-		close(es.byDaysCh)
-		close(es.byNewerNoncurrentCh)
+		xioutil.SafeClose(es.byDaysCh)
+		xioutil.SafeClose(es.byNewerNoncurrentCh)
 	})
 }
 
@@ -217,17 +220,24 @@ type transitionState struct {
 	numWorkers int
 	killCh     chan struct{}
 
-	activeTasks int32
+	activeTasks          atomic.Int64
+	missedImmediateTasks atomic.Int64
 
 	lastDayMu    sync.RWMutex
 	lastDayStats map[string]*lastDayTierStats
 }
 
 func (t *transitionState) queueTransitionTask(oi ObjectInfo, event lifecycle.Event, src lcEventSrc) {
+	task := transitionTask{objInfo: oi, event: event, src: src}
 	select {
 	case <-t.ctx.Done():
-	case t.transitionCh <- transitionTask{objInfo: oi, event: event, src: src}:
+	case t.transitionCh <- task:
 	default:
+		switch src {
+		case lcEventSrc_s3PutObject, lcEventSrc_s3CopyObject, lcEventSrc_s3CompleteMultipartUpload:
+			// Update missed immediate tasks only for incoming requests.
+			t.missedImmediateTasks.Add(1)
+		}
 	}
 }
 
@@ -237,7 +247,7 @@ var globalTransitionState *transitionState
 // via its Init method.
 func newTransitionState(ctx context.Context) *transitionState {
 	return &transitionState{
-		transitionCh: make(chan transitionTask, 10000),
+		transitionCh: make(chan transitionTask, 100000),
 		ctx:          ctx,
 		killCh:       make(chan struct{}),
 		lastDayStats: make(map[string]*lastDayTierStats),
@@ -262,8 +272,14 @@ func (t *transitionState) PendingTasks() int {
 }
 
 // ActiveTasks returns the number of active (ongoing) ILM transition tasks.
-func (t *transitionState) ActiveTasks() int {
-	return int(atomic.LoadInt32(&t.activeTasks))
+func (t *transitionState) ActiveTasks() int64 {
+	return t.activeTasks.Load()
+}
+
+// MissedImmediateTasks returns the number of tasks - deferred to scanner due
+// to tasks channel being backlogged.
+func (t *transitionState) MissedImmediateTasks() int64 {
+	return t.missedImmediateTasks.Load()
 }
 
 // worker waits for transition tasks
@@ -278,11 +294,13 @@ func (t *transitionState) worker(objectAPI ObjectLayer) {
 			if !ok {
 				return
 			}
-			atomic.AddInt32(&t.activeTasks, 1)
+			t.activeTasks.Add(1)
 			if err := transitionObject(t.ctx, objectAPI, task.objInfo, newLifecycleAuditEvent(task.src, task.event)); err != nil {
-				if !isErrVersionNotFound(err) && !isErrObjectNotFound(err) {
-					logger.LogIf(t.ctx, fmt.Errorf("Transition to %s failed for %s/%s version:%s with %w",
-						task.event.StorageClass, task.objInfo.Bucket, task.objInfo.Name, task.objInfo.VersionID, err))
+				if !isErrVersionNotFound(err) && !isErrObjectNotFound(err) && !xnet.IsNetworkOrHostDown(err, false) {
+					if !strings.Contains(err.Error(), "use of closed network connection") {
+						logger.LogIf(t.ctx, fmt.Errorf("Transition to %s failed for %s/%s version:%s with %w",
+							task.event.StorageClass, task.objInfo.Bucket, task.objInfo.Name, task.objInfo.VersionID, err))
+					}
 				}
 			} else {
 				ts := tierStats{
@@ -294,7 +312,7 @@ func (t *transitionState) worker(objectAPI ObjectLayer) {
 				}
 				t.addLastDayStats(task.event.StorageClass, ts)
 			}
-			atomic.AddInt32(&t.activeTasks, -1)
+			t.activeTasks.Add(-1)
 		}
 	}
 }
@@ -444,7 +462,8 @@ func genTransitionObjName(bucket string) (string, error) {
 		return "", err
 	}
 	us := u.String()
-	obj := fmt.Sprintf("%s/%s/%s/%s/%s", globalDeploymentID, bucket, us[0:2], us[2:4], us)
+	hash := xxh3.HashString(pathJoin(globalDeploymentID(), bucket))
+	obj := fmt.Sprintf("%s/%s/%s/%s", strconv.FormatUint(hash, 16), us[0:2], us[2:4], us)
 	return obj, nil
 }
 
@@ -489,9 +508,13 @@ func auditTierActions(ctx context.Context, tier string, bytes int64) func(err er
 		}
 
 		if err == nil {
-			op.TimeToResponseNS = time.Since(startTime).Nanoseconds()
+			since := time.Since(startTime)
+			op.TimeToResponseNS = since.Nanoseconds()
+			globalTierMetrics.Observe(tier, since)
+			globalTierMetrics.logSuccess(tier)
 		} else {
 			op.Error = err.Error()
+			globalTierMetrics.logFailure(tier)
 		}
 
 		logger.GetReqInfo(ctx).AppendTags("tierStats", op)
@@ -862,6 +885,7 @@ func (oi ObjectInfo) ToLifecycleOpts() lifecycle.ObjectOpts {
 		UserTags:         oi.UserTags,
 		VersionID:        oi.VersionID,
 		ModTime:          oi.ModTime,
+		Size:             oi.Size,
 		IsLatest:         oi.IsLatest,
 		NumVersions:      oi.NumVersions,
 		DeleteMarker:     oi.DeleteMarker,

cmd/bucket-listobjects-handlers.go

@@ -111,7 +111,7 @@ func (api objectAPIHandlers) listObjectVersionsHandler(w http.ResponseWriter, r
 
 	listObjectVersions := objectAPI.ListObjectVersions
 
-	// Inititate a list object versions operation based on the input params.
+	// Initiate a list object versions operation based on the input params.
 	// On success would return back ListObjectsInfo object to be
 	// marshaled into S3 compatible XML header.
 	listObjectVersionsInfo, err := listObjectVersions(ctx, bucket, prefix, marker, versionIDMarker, delimiter, maxkeys)
@@ -201,10 +201,10 @@ func (api objectAPIHandlers) listObjectsV2Handler(ctx context.Context, w http.Re
 	)
 
 	if r.Header.Get(xMinIOExtract) == "true" && strings.Contains(prefix, archivePattern) {
-		// Inititate a list objects operation inside a zip file based in the input params
+		// Initiate a list objects operation inside a zip file based in the input params
 		listObjectsV2Info, err = listObjectsV2InArchive(ctx, objectAPI, bucket, prefix, token, delimiter, maxKeys, fetchOwner, startAfter)
 	} else {
-		// Inititate a list objects operation based on the input params.
+		// Initiate a list objects operation based on the input params.
 		// On success would return back ListObjectsInfo object to be
 		// marshaled into S3 compatible XML header.
 		listObjectsV2Info, err = objectAPI.ListObjectsV2(ctx, bucket, prefix, token, delimiter, maxKeys, fetchOwner, startAfter)
@@ -304,7 +304,7 @@ func (api objectAPIHandlers) ListObjectsV1Handler(w http.ResponseWriter, r *http
 
 	listObjects := objectAPI.ListObjects
 
-	// Inititate a list objects operation based on the input params.
+	// Initiate a list objects operation based on the input params.
 	// On success would return back ListObjectsInfo object to be
 	// marshaled into S3 compatible XML header.
 	listObjectsInfo, err := listObjects(ctx, bucket, prefix, marker, delimiter, maxKeys)

cmd/bucket-metadata-sys.go

@@ -19,8 +19,10 @@ package cmd
 
 import (
 	"context"
+	"encoding/xml"
 	"errors"
 	"fmt"
+	"math/rand"
 	"sync"
 	"time"
 
@@ -177,6 +179,40 @@ func (sys *BucketMetadataSys) save(ctx context.Context, meta BucketMetadata) err
 // Delete delete the bucket metadata for the specified bucket.
 // must be used by all callers instead of using Update() with nil configData.
 func (sys *BucketMetadataSys) Delete(ctx context.Context, bucket string, configFile string) (updatedAt time.Time, err error) {
+	if configFile == bucketLifecycleConfig {
+		// Get bucket config from current site
+		meta, e := globalBucketMetadataSys.GetConfigFromDisk(ctx, bucket)
+		if e != nil && !errors.Is(e, errConfigNotFound) {
+			return updatedAt, e
+		}
+		var expiryRuleRemoved bool
+		if len(meta.LifecycleConfigXML) > 0 {
+			var lcCfg lifecycle.Lifecycle
+			if err := xml.Unmarshal(meta.LifecycleConfigXML, &lcCfg); err != nil {
+				return updatedAt, err
+			}
+			// find a single expiry rule set the flag
+			for _, rl := range lcCfg.Rules {
+				if !rl.Expiration.IsNull() || !rl.NoncurrentVersionExpiration.IsNull() {
+					expiryRuleRemoved = true
+					break
+				}
+			}
+		}
+
+		// Form empty ILM details with `ExpiryUpdatedAt` field and save
+		var cfgData []byte
+		if expiryRuleRemoved {
+			var lcCfg lifecycle.Lifecycle
+			currtime := time.Now()
+			lcCfg.ExpiryUpdatedAt = &currtime
+			cfgData, err = xml.Marshal(lcCfg)
+			if err != nil {
+				return updatedAt, err
+			}
+		}
+		return sys.updateAndParse(ctx, bucket, configFile, cfgData, false)
+	}
 	return sys.updateAndParse(ctx, bucket, configFile, nil, false)
 }
 
@@ -267,7 +303,10 @@ func (sys *BucketMetadataSys) GetLifecycleConfig(bucket string) (*lifecycle.Life
 		}
 		return nil, time.Time{}, err
 	}
-	if meta.lifecycleConfig == nil {
+	// there could be just `ExpiryUpdatedAt` field populated as part
+	// of last delete all. Treat this situation as not lifecycle configuration
+	// available
+	if meta.lifecycleConfig == nil || len(meta.lifecycleConfig.Rules) == 0 {
 		return nil, time.Time{}, BucketLifecycleNotFound{Bucket: bucket}
 	}
 	return meta.lifecycleConfig, meta.LifecycleConfigUpdatedAt, nil
@@ -437,31 +476,18 @@ func (sys *BucketMetadataSys) Init(ctx context.Context, buckets []BucketInfo, ob
 	return nil
 }
 
-func (sys *BucketMetadataSys) loadBucketMetadata(ctx context.Context, bucket BucketInfo) error {
-	meta, err := loadBucketMetadata(ctx, sys.objAPI, bucket.Name)
-	if err != nil {
-		return err
-	}
-
-	sys.Lock()
-	sys.metadataMap[bucket.Name] = meta
-	sys.Unlock()
-
-	return nil
-}
-
 // concurrently load bucket metadata to speed up loading bucket metadata.
-func (sys *BucketMetadataSys) concurrentLoad(ctx context.Context, buckets []BucketInfo) {
+func (sys *BucketMetadataSys) concurrentLoad(ctx context.Context, buckets []BucketInfo, failedBuckets map[string]struct{}) {
 	g := errgroup.WithNErrs(len(buckets))
 	bucketMetas := make([]BucketMetadata, len(buckets))
 	for index := range buckets {
 		index := index
 		g.Go(func() error {
-			_, _ = sys.objAPI.HealBucket(ctx, buckets[index].Name, madmin.HealOpts{
-				// Ensure heal opts for bucket metadata be deep healed all the time.
-				ScanMode: madmin.HealDeepScan,
-				Recreate: true,
-			})
+			// Sleep and stagger to avoid blocked CPU and thundering
+			// herd upon start up sequence.
+			time.Sleep(25*time.Millisecond + time.Duration(rand.Int63n(int64(100*time.Millisecond))))
+
+			_, _ = sys.objAPI.HealBucket(ctx, buckets[index].Name, madmin.HealOpts{Recreate: true})
 			meta, err := loadBucketMetadata(ctx, sys.objAPI, buckets[index].Name)
 			if err != nil {
 				return err
@@ -491,6 +517,10 @@ func (sys *BucketMetadataSys) concurrentLoad(ctx context.Context, buckets []Buck
 
 	for i, meta := range bucketMetas {
 		if errs[i] != nil {
+			if failedBuckets == nil {
+				failedBuckets = make(map[string]struct{})
+			}
+			failedBuckets[buckets[i].Name] = struct{}{}
 			continue
 		}
 		globalEventNotifier.set(buckets[i], meta)   // set notification targets
@@ -498,7 +528,7 @@ func (sys *BucketMetadataSys) concurrentLoad(ctx context.Context, buckets []Buck
 	}
 }
 
-func (sys *BucketMetadataSys) refreshBucketsMetadataLoop(ctx context.Context) {
+func (sys *BucketMetadataSys) refreshBucketsMetadataLoop(ctx context.Context, failedBuckets map[string]struct{}) {
 	const bucketMetadataRefresh = 15 * time.Minute
 
 	sleeper := newDynamicSleeper(2, 150*time.Millisecond, false)
@@ -513,7 +543,7 @@ func (sys *BucketMetadataSys) refreshBucketsMetadataLoop(ctx context.Context) {
 			buckets, err := sys.objAPI.ListBuckets(ctx, BucketOptions{})
 			if err != nil {
 				logger.LogIf(ctx, err)
-				continue
+				break
 			}
 
 			// Handle if we have some buckets in-memory those are stale.
@@ -525,38 +555,49 @@ func (sys *BucketMetadataSys) refreshBucketsMetadataLoop(ctx context.Context) {
 			}
 			sys.RemoveStaleBuckets(diskBuckets)
 
-			for _, bucket := range buckets {
+			for i := range buckets {
 				wait := sleeper.Timer(ctx)
 
-				err := sys.loadBucketMetadata(ctx, bucket)
+				meta, err := loadBucketMetadata(ctx, sys.objAPI, buckets[i].Name)
 				if err != nil {
 					logger.LogIf(ctx, err)
 					wait() // wait to proceed to next entry.
 					continue
 				}
 
+				sys.Lock()
+				sys.metadataMap[buckets[i].Name] = meta
+				sys.Unlock()
+
+				// Initialize the failed buckets
+				if _, ok := failedBuckets[buckets[i].Name]; ok {
+					globalEventNotifier.set(buckets[i], meta)
+					globalBucketTargetSys.set(buckets[i], meta)
+					delete(failedBuckets, buckets[i].Name)
+				}
+
 				wait() // wait to proceed to next entry.
 			}
-
-			t.Reset(bucketMetadataRefresh)
 		}
+		t.Reset(bucketMetadataRefresh)
 	}
 }
 
 // Loads bucket metadata for all buckets into BucketMetadataSys.
 func (sys *BucketMetadataSys) init(ctx context.Context, buckets []BucketInfo) {
 	count := 100 // load 100 bucket metadata at a time.
+	failedBuckets := make(map[string]struct{})
 	for {
 		if len(buckets) < count {
-			sys.concurrentLoad(ctx, buckets)
+			sys.concurrentLoad(ctx, buckets, failedBuckets)
 			break
 		}
-		sys.concurrentLoad(ctx, buckets[:count])
+		sys.concurrentLoad(ctx, buckets[:count], failedBuckets)
 		buckets = buckets[count:]
 	}
 
 	if globalIsDistErasure {
-		go sys.refreshBucketsMetadataLoop(ctx)
+		go sys.refreshBucketsMetadataLoop(ctx, failedBuckets)
 	}
 }
 

cmd/bucket-metadata.go

@@ -182,26 +182,34 @@ func loadBucketMetadataParse(ctx context.Context, objectAPI ObjectLayer, bucket
 		b.defaultTimestamps()
 	}
 
-	configs, err := b.getAllLegacyConfigs(ctx, objectAPI)
-	if err != nil {
-		return b, err
+	// If bucket metadata is missing look for legacy files,
+	// since we only ever had b.Created as non-zero when
+	// migration was complete in 2020-May release. So this
+	// a check to avoid migrating for buckets that already
+	// have this field set.
+	if b.Created.IsZero() {
+		configs, err := b.getAllLegacyConfigs(ctx, objectAPI)
+		if err != nil {
+			return b, err
+		}
+
+		if len(configs) > 0 {
+			// Old bucket without bucket metadata. Hence we migrate existing settings.
+			if err = b.convertLegacyConfigs(ctx, objectAPI, configs); err != nil {
+				return b, err
+			}
+		}
 	}
 
-	if len(configs) == 0 {
-		if parse {
-			// nothing to update, parse and proceed.
-			err = b.parseAllConfigs(ctx, objectAPI)
+	if parse {
+		// nothing to update, parse and proceed.
+		if err = b.parseAllConfigs(ctx, objectAPI); err != nil {
+			return b, err
 		}
-	} else {
-		// Old bucket without bucket metadata. Hence we migrate existing settings.
-		err = b.convertLegacyConfigs(ctx, objectAPI, configs)
-	}
-	if err != nil {
-		return b, err
 	}
 
 	// migrate unencrypted remote targets
-	if err := b.migrateTargetConfig(ctx, objectAPI); err != nil {
+	if err = b.migrateTargetConfig(ctx, objectAPI); err != nil {
 		return b, err
 	}
 
@@ -331,7 +339,7 @@ func (b *BucketMetadata) getAllLegacyConfigs(ctx context.Context, objectAPI Obje
 	for _, legacyFile := range legacyConfigs {
 		configFile := path.Join(bucketMetaPrefix, b.Name, legacyFile)
 
-		configData, err := readConfig(ctx, objectAPI, configFile)
+		configData, info, err := readConfigWithMetadata(ctx, objectAPI, configFile, ObjectOptions{})
 		if err != nil {
 			if _, ok := err.(ObjectExistsAsDirectory); ok {
 				// in FS mode it possible that we have actual
@@ -346,6 +354,7 @@ func (b *BucketMetadata) getAllLegacyConfigs(ctx context.Context, objectAPI Obje
 			return nil, err
 		}
 		configs[legacyFile] = configData
+		b.Created = info.ModTime
 	}
 
 	return configs, nil

cmd/bucket-policy-handlers_test.go

@@ -317,7 +317,7 @@ func testPutBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string
 		if err != nil {
 			t.Fatalf("Test %d: %s: Failed to create HTTP request for PutBucketPolicyHandler: <ERROR> %v", i+1, instanceType, err)
 		}
-		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic ofthe handler.
+		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic of the handler.
 		// Call the ServeHTTP to execute the handler.
 		apiRouter.ServeHTTP(recV4, reqV4)
 		if recV4.Code != testCase.expectedRespStatus {
@@ -331,7 +331,7 @@ func testPutBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string
 		if err != nil {
 			t.Fatalf("Test %d: %s: Failed to create HTTP request for PutBucketPolicyHandler: <ERROR> %v", i+1, instanceType, err)
 		}
-		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic ofthe handler.
+		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic of the handler.
 		// Call the ServeHTTP to execute the handler.
 		apiRouter.ServeHTTP(recV2, reqV2)
 		if recV2.Code != testCase.expectedRespStatus {
@@ -407,7 +407,7 @@ func testGetBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string
 		if err != nil {
 			t.Fatalf("Test %d: Failed to create HTTP request for PutBucketPolicyHandler: <ERROR> %v", i+1, err)
 		}
-		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic ofthe handler.
+		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic of the handler.
 		// Call the ServeHTTP to execute the handler.
 		apiRouter.ServeHTTP(recV4, reqV4)
 		if recV4.Code != testPolicy.expectedRespStatus {
@@ -421,7 +421,7 @@ func testGetBucketPolicyHandler(obj ObjectLayer, instanceType, bucketName string
 		if err != nil {
 			t.Fatalf("Test %d: Failed to create HTTP request for PutBucketPolicyHandler: <ERROR> %v", i+1, err)
 		}
-		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic ofthe handler.
+		// Since `apiRouter` satisfies `http.Handler` it has a ServeHTTP to execute the logic of the handler.
 		// Call the ServeHTTP to execute the handler.
 		apiRouter.ServeHTTP(recV2, reqV2)
 		if recV2.Code != testPolicy.expectedRespStatus {

cmd/bucket-policy.go

@@ -66,6 +66,14 @@ func NewPolicySys() *PolicySys {
 	return &PolicySys{}
 }
 
+func getSTSConditionValues(r *http.Request, lc string, cred auth.Credentials) map[string][]string {
+	m := make(map[string][]string)
+	if d := r.Form.Get("DurationSeconds"); d != "" {
+		m["DurationSeconds"] = []string{d}
+	}
+	return m
+}
+
 func getConditionValues(r *http.Request, lc string, cred auth.Credentials) map[string][]string {
 	currTime := UTCNow()
 
@@ -121,7 +129,7 @@ func getConditionValues(r *http.Request, lc string, cred auth.Credentials) map[s
 		"CurrentTime":      {currTime.Format(time.RFC3339)},
 		"EpochTime":        {strconv.FormatInt(currTime.Unix(), 10)},
 		"SecureTransport":  {strconv.FormatBool(r.TLS != nil)},
-		"SourceIp":         {handlers.GetSourceIP(r)},
+		"SourceIp":         {handlers.GetSourceIPRaw(r)},
 		"UserAgent":        {r.UserAgent()},
 		"Referer":          {r.Referer()},
 		"principaltype":    {principalType},
@@ -137,6 +145,10 @@ func getConditionValues(r *http.Request, lc string, cred auth.Credentials) map[s
 	}
 
 	cloneHeader := r.Header.Clone()
+	if v := cloneHeader.Get("x-amz-signature-age"); v != "" {
+		args["signatureAge"] = []string{v}
+		cloneHeader.Del("x-amz-signature-age")
+	}
 
 	if userTags := cloneHeader.Get(xhttp.AmzObjectTagging); userTags != "" {
 		tag, _ := tags.ParseObjectTags(userTags)

cmd/bucket-replication-handlers.go

@@ -575,7 +575,7 @@ func (api objectAPIHandlers) ValidateBucketReplicationCredsHandler(w http.Respon
 		if rule.Status == replication.Disabled {
 			continue
 		}
-		clnt := globalBucketTargetSys.GetRemoteTargetClient(rule.Destination.Bucket)
+		clnt := globalBucketTargetSys.GetRemoteTargetClient(bucket, rule.Destination.Bucket)
 		if clnt == nil {
 			writeErrorResponse(ctx, w, errorCodes.ToAPIErrWithErr(ErrRemoteTargetNotFoundError, fmt.Errorf("replication config with rule ID %s has a stale target", rule.ID)), r.URL)
 			return

cmd/bucket-replication-metrics.go

@@ -337,13 +337,13 @@ type SMA struct {
 	filledBuf bool
 }
 
-func newSMA(len int) *SMA {
-	if len <= 0 {
-		len = defaultWindowSize
+func newSMA(ln int) *SMA {
+	if ln <= 0 {
+		ln = defaultWindowSize
 	}
 	return &SMA{
-		buf:    make([]float64, len),
-		window: len,
+		buf:    make([]float64, ln),
+		window: ln,
 		idx:    0,
 	}
 }
@@ -379,3 +379,145 @@ func (s *SMA) simpleMovingAvg() float64 {
 const (
 	defaultWindowSize = 10
 )
+
+type proxyStatsCache struct {
+	srProxyStats ProxyMetric
+	bucketStats  map[string]ProxyMetric
+	sync.RWMutex // mutex for proxy stats
+}
+
+func newProxyStatsCache() proxyStatsCache {
+	return proxyStatsCache{
+		bucketStats: make(map[string]ProxyMetric),
+	}
+}
+
+func (p *proxyStatsCache) inc(bucket string, api replProxyAPI, isErr bool) {
+	p.Lock()
+	defer p.Unlock()
+	v, ok := p.bucketStats[bucket]
+	if !ok {
+		v = ProxyMetric{}
+	}
+	switch api {
+	case putObjectTaggingAPI:
+		if !isErr {
+			atomic.AddUint64(&v.PutTagTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.PutTagTotal, 1)
+		} else {
+			atomic.AddUint64(&v.PutTagFailedTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.PutTagFailedTotal, 1)
+		}
+	case getObjectTaggingAPI:
+		if !isErr {
+			atomic.AddUint64(&v.GetTagTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.GetTagTotal, 1)
+		} else {
+			atomic.AddUint64(&v.GetTagFailedTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.GetTagFailedTotal, 1)
+		}
+	case removeObjectTaggingAPI:
+		if !isErr {
+			atomic.AddUint64(&v.RmvTagTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.RmvTagTotal, 1)
+		} else {
+			atomic.AddUint64(&v.RmvTagFailedTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.RmvTagFailedTotal, 1)
+		}
+	case headObjectAPI:
+		if !isErr {
+			atomic.AddUint64(&v.HeadTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.HeadTotal, 1)
+		} else {
+			atomic.AddUint64(&v.HeadFailedTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.HeadFailedTotal, 1)
+		}
+	case getObjectAPI:
+		if !isErr {
+			atomic.AddUint64(&v.GetTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.GetTotal, 1)
+		} else {
+			atomic.AddUint64(&v.GetFailedTotal, 1)
+			atomic.AddUint64(&p.srProxyStats.GetFailedTotal, 1)
+		}
+	default:
+		return
+	}
+	p.bucketStats[bucket] = v
+}
+
+func (p *proxyStatsCache) getBucketStats(bucket string) ProxyMetric {
+	p.RLock()
+	defer p.RUnlock()
+	v, ok := p.bucketStats[bucket]
+
+	if !ok {
+		return ProxyMetric{}
+	}
+	return ProxyMetric{
+		PutTagTotal: atomic.LoadUint64(&v.PutTagTotal),
+		GetTagTotal: atomic.LoadUint64(&v.GetTagTotal),
+		RmvTagTotal: atomic.LoadUint64(&v.RmvTagTotal),
+		HeadTotal:   atomic.LoadUint64(&v.HeadTotal),
+		GetTotal:    atomic.LoadUint64(&v.GetTotal),
+
+		PutTagFailedTotal: atomic.LoadUint64(&v.PutTagFailedTotal),
+		GetTagFailedTotal: atomic.LoadUint64(&v.GetTagFailedTotal),
+		RmvTagFailedTotal: atomic.LoadUint64(&v.RmvTagFailedTotal),
+		HeadFailedTotal:   atomic.LoadUint64(&v.HeadFailedTotal),
+		GetFailedTotal:    atomic.LoadUint64(&v.GetFailedTotal),
+	}
+}
+
+func (p *proxyStatsCache) getSiteStats() ProxyMetric {
+	v := p.srProxyStats
+	return ProxyMetric{
+		PutTagTotal:       atomic.LoadUint64(&v.PutTagTotal),
+		GetTagTotal:       atomic.LoadUint64(&v.GetTagTotal),
+		RmvTagTotal:       atomic.LoadUint64(&v.RmvTagTotal),
+		HeadTotal:         atomic.LoadUint64(&v.HeadTotal),
+		GetTotal:          atomic.LoadUint64(&v.GetTotal),
+		PutTagFailedTotal: atomic.LoadUint64(&v.PutTagFailedTotal),
+		GetTagFailedTotal: atomic.LoadUint64(&v.GetTagFailedTotal),
+		RmvTagFailedTotal: atomic.LoadUint64(&v.RmvTagFailedTotal),
+		HeadFailedTotal:   atomic.LoadUint64(&v.HeadFailedTotal),
+		GetFailedTotal:    atomic.LoadUint64(&v.GetFailedTotal),
+	}
+}
+
+type replProxyAPI string
+
+const (
+	putObjectTaggingAPI    replProxyAPI = "PutObjectTagging"
+	getObjectTaggingAPI    replProxyAPI = "GetObjectTagging"
+	removeObjectTaggingAPI replProxyAPI = "RemoveObjectTagging"
+	headObjectAPI          replProxyAPI = "HeadObject"
+	getObjectAPI           replProxyAPI = "GetObject"
+)
+
+// ProxyMetric holds stats for replication proxying
+type ProxyMetric struct {
+	PutTagTotal       uint64 `json:"putTaggingProxyTotal" msg:"ptc"`
+	GetTagTotal       uint64 `json:"getTaggingProxyTotal" msg:"gtc"`
+	RmvTagTotal       uint64 `json:"removeTaggingProxyTotal" msg:"rtc"`
+	GetTotal          uint64 `json:"getProxyTotal" msg:"gc"`
+	HeadTotal         uint64 `json:"headProxyTotal" msg:"hc"`
+	PutTagFailedTotal uint64 `json:"putTaggingProxyFailed" msg:"ptf"`
+	GetTagFailedTotal uint64 `json:"getTaggingProxyFailed" msg:"gtf"`
+	RmvTagFailedTotal uint64 `json:"removeTaggingProxyFailed" msg:"rtf"`
+	GetFailedTotal    uint64 `json:"getProxyFailed" msg:"gf"`
+	HeadFailedTotal   uint64 `json:"headProxyFailed" msg:"hf"`
+}
+
+func (p *ProxyMetric) add(p2 ProxyMetric) {
+	atomic.AddUint64(&p.GetTotal, p2.GetTotal)
+	atomic.AddUint64(&p.HeadTotal, p2.HeadTotal)
+	atomic.AddUint64(&p.GetTagTotal, p2.GetTagTotal)
+	atomic.AddUint64(&p.PutTagTotal, p2.PutTagTotal)
+	atomic.AddUint64(&p.RmvTagTotal, p2.RmvTagTotal)
+	atomic.AddUint64(&p.GetFailedTotal, p2.GetFailedTotal)
+	atomic.AddUint64(&p.HeadFailedTotal, p2.HeadFailedTotal)
+	atomic.AddUint64(&p.GetTagFailedTotal, p2.GetTagFailedTotal)
+	atomic.AddUint64(&p.PutTagFailedTotal, p2.PutTagFailedTotal)
+	atomic.AddUint64(&p.RmvTagFailedTotal, p2.RmvTagFailedTotal)
+}

cmd/bucket-replication-metrics_gen.go

@@ -635,6 +635,334 @@ func (z InQueueStats) Msgsize() (s int) {
 	return
 }
 
+// DecodeMsg implements msgp.Decodable
+func (z *ProxyMetric) DecodeMsg(dc *msgp.Reader) (err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, err = dc.ReadMapHeader()
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, err = dc.ReadMapKeyPtr()
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "ptc":
+			z.PutTagTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "PutTagTotal")
+				return
+			}
+		case "gtc":
+			z.GetTagTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "GetTagTotal")
+				return
+			}
+		case "rtc":
+			z.RmvTagTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "RmvTagTotal")
+				return
+			}
+		case "gc":
+			z.GetTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "GetTotal")
+				return
+			}
+		case "hc":
+			z.HeadTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "HeadTotal")
+				return
+			}
+		case "ptf":
+			z.PutTagFailedTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "PutTagFailedTotal")
+				return
+			}
+		case "gtf":
+			z.GetTagFailedTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "GetTagFailedTotal")
+				return
+			}
+		case "rtf":
+			z.RmvTagFailedTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "RmvTagFailedTotal")
+				return
+			}
+		case "gf":
+			z.GetFailedTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "GetFailedTotal")
+				return
+			}
+		case "hf":
+			z.HeadFailedTotal, err = dc.ReadUint64()
+			if err != nil {
+				err = msgp.WrapError(err, "HeadFailedTotal")
+				return
+			}
+		default:
+			err = dc.Skip()
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	return
+}
+
+// EncodeMsg implements msgp.Encodable
+func (z *ProxyMetric) EncodeMsg(en *msgp.Writer) (err error) {
+	// map header, size 10
+	// write "ptc"
+	err = en.Append(0x8a, 0xa3, 0x70, 0x74, 0x63)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.PutTagTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "PutTagTotal")
+		return
+	}
+	// write "gtc"
+	err = en.Append(0xa3, 0x67, 0x74, 0x63)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.GetTagTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "GetTagTotal")
+		return
+	}
+	// write "rtc"
+	err = en.Append(0xa3, 0x72, 0x74, 0x63)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.RmvTagTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "RmvTagTotal")
+		return
+	}
+	// write "gc"
+	err = en.Append(0xa2, 0x67, 0x63)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.GetTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "GetTotal")
+		return
+	}
+	// write "hc"
+	err = en.Append(0xa2, 0x68, 0x63)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.HeadTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "HeadTotal")
+		return
+	}
+	// write "ptf"
+	err = en.Append(0xa3, 0x70, 0x74, 0x66)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.PutTagFailedTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "PutTagFailedTotal")
+		return
+	}
+	// write "gtf"
+	err = en.Append(0xa3, 0x67, 0x74, 0x66)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.GetTagFailedTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "GetTagFailedTotal")
+		return
+	}
+	// write "rtf"
+	err = en.Append(0xa3, 0x72, 0x74, 0x66)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.RmvTagFailedTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "RmvTagFailedTotal")
+		return
+	}
+	// write "gf"
+	err = en.Append(0xa2, 0x67, 0x66)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.GetFailedTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "GetFailedTotal")
+		return
+	}
+	// write "hf"
+	err = en.Append(0xa2, 0x68, 0x66)
+	if err != nil {
+		return
+	}
+	err = en.WriteUint64(z.HeadFailedTotal)
+	if err != nil {
+		err = msgp.WrapError(err, "HeadFailedTotal")
+		return
+	}
+	return
+}
+
+// MarshalMsg implements msgp.Marshaler
+func (z *ProxyMetric) MarshalMsg(b []byte) (o []byte, err error) {
+	o = msgp.Require(b, z.Msgsize())
+	// map header, size 10
+	// string "ptc"
+	o = append(o, 0x8a, 0xa3, 0x70, 0x74, 0x63)
+	o = msgp.AppendUint64(o, z.PutTagTotal)
+	// string "gtc"
+	o = append(o, 0xa3, 0x67, 0x74, 0x63)
+	o = msgp.AppendUint64(o, z.GetTagTotal)
+	// string "rtc"
+	o = append(o, 0xa3, 0x72, 0x74, 0x63)
+	o = msgp.AppendUint64(o, z.RmvTagTotal)
+	// string "gc"
+	o = append(o, 0xa2, 0x67, 0x63)
+	o = msgp.AppendUint64(o, z.GetTotal)
+	// string "hc"
+	o = append(o, 0xa2, 0x68, 0x63)
+	o = msgp.AppendUint64(o, z.HeadTotal)
+	// string "ptf"
+	o = append(o, 0xa3, 0x70, 0x74, 0x66)
+	o = msgp.AppendUint64(o, z.PutTagFailedTotal)
+	// string "gtf"
+	o = append(o, 0xa3, 0x67, 0x74, 0x66)
+	o = msgp.AppendUint64(o, z.GetTagFailedTotal)
+	// string "rtf"
+	o = append(o, 0xa3, 0x72, 0x74, 0x66)
+	o = msgp.AppendUint64(o, z.RmvTagFailedTotal)
+	// string "gf"
+	o = append(o, 0xa2, 0x67, 0x66)
+	o = msgp.AppendUint64(o, z.GetFailedTotal)
+	// string "hf"
+	o = append(o, 0xa2, 0x68, 0x66)
+	o = msgp.AppendUint64(o, z.HeadFailedTotal)
+	return
+}
+
+// UnmarshalMsg implements msgp.Unmarshaler
+func (z *ProxyMetric) UnmarshalMsg(bts []byte) (o []byte, err error) {
+	var field []byte
+	_ = field
+	var zb0001 uint32
+	zb0001, bts, err = msgp.ReadMapHeaderBytes(bts)
+	if err != nil {
+		err = msgp.WrapError(err)
+		return
+	}
+	for zb0001 > 0 {
+		zb0001--
+		field, bts, err = msgp.ReadMapKeyZC(bts)
+		if err != nil {
+			err = msgp.WrapError(err)
+			return
+		}
+		switch msgp.UnsafeString(field) {
+		case "ptc":
+			z.PutTagTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "PutTagTotal")
+				return
+			}
+		case "gtc":
+			z.GetTagTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "GetTagTotal")
+				return
+			}
+		case "rtc":
+			z.RmvTagTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "RmvTagTotal")
+				return
+			}
+		case "gc":
+			z.GetTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "GetTotal")
+				return
+			}
+		case "hc":
+			z.HeadTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "HeadTotal")
+				return
+			}
+		case "ptf":
+			z.PutTagFailedTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "PutTagFailedTotal")
+				return
+			}
+		case "gtf":
+			z.GetTagFailedTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "GetTagFailedTotal")
+				return
+			}
+		case "rtf":
+			z.RmvTagFailedTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "RmvTagFailedTotal")
+				return
+			}
+		case "gf":
+			z.GetFailedTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "GetFailedTotal")
+				return
+			}
+		case "hf":
+			z.HeadFailedTotal, bts, err = msgp.ReadUint64Bytes(bts)
+			if err != nil {
+				err = msgp.WrapError(err, "HeadFailedTotal")
+				return
+			}
+		default:
+			bts, err = msgp.Skip(bts)
+			if err != nil {
+				err = msgp.WrapError(err)
+				return
+			}
+		}
+	}
+	o = bts
+	return
+}
+
+// Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message
+func (z *ProxyMetric) Msgsize() (s int) {
+	s = 1 + 4 + msgp.Uint64Size + 4 + msgp.Uint64Size + 4 + msgp.Uint64Size + 3 + msgp.Uint64Size + 3 + msgp.Uint64Size + 4 + msgp.Uint64Size + 4 + msgp.Uint64Size + 4 + msgp.Uint64Size + 3 + msgp.Uint64Size + 3 + msgp.Uint64Size
+	return
+}
+
 // DecodeMsg implements msgp.Decodable
 func (z *QStat) DecodeMsg(dc *msgp.Reader) (err error) {
 	var field []byte

cmd/bucket-replication-metrics_gen_test.go

@@ -348,6 +348,119 @@ func BenchmarkDecodeInQueueStats(b *testing.B) {
 	}
 }
 
+func TestMarshalUnmarshalProxyMetric(t *testing.T) {
+	v := ProxyMetric{}
+	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 BenchmarkMarshalMsgProxyMetric(b *testing.B) {
+	v := ProxyMetric{}
+	b.ReportAllocs()
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		v.MarshalMsg(nil)
+	}
+}
+
+func BenchmarkAppendMsgProxyMetric(b *testing.B) {
+	v := ProxyMetric{}
+	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 BenchmarkUnmarshalProxyMetric(b *testing.B) {
+	v := ProxyMetric{}
+	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 TestEncodeDecodeProxyMetric(t *testing.T) {
+	v := ProxyMetric{}
+	var buf bytes.Buffer
+	msgp.Encode(&buf, &v)
+
+	m := v.Msgsize()
+	if buf.Len() > m {
+		t.Log("WARNING: TestEncodeDecodeProxyMetric Msgsize() is inaccurate")
+	}
+
+	vn := ProxyMetric{}
+	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 BenchmarkEncodeProxyMetric(b *testing.B) {
+	v := ProxyMetric{}
+	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 BenchmarkDecodeProxyMetric(b *testing.B) {
+	v := ProxyMetric{}
+	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 TestMarshalUnmarshalQStat(t *testing.T) {
 	v := QStat{}
 	bts, err := v.MarshalMsg(nil)

cmd/bucket-replication-stats.go

@@ -45,6 +45,8 @@ type ReplicationStats struct {
 	workers *ActiveWorkerStat
 	// queue stats cache
 	qCache queueCache
+
+	pCache proxyStatsCache
 	// mrf backlog stats
 	mrfStats ReplicationMRFStats
 	// for bucket replication, continue to use existing cache
@@ -305,6 +307,7 @@ func (r *ReplicationStats) getSRMetricsForNode() SRMetricsSummary {
 		Queued:        r.qCache.getSiteStats(),
 		ActiveWorkers: r.ActiveWorkers(),
 		Metrics:       r.srStats.get(),
+		Proxied:       r.pCache.getSiteStats(),
 		ReplicaSize:   atomic.LoadInt64(&r.srStats.ReplicaSize),
 		ReplicaCount:  atomic.LoadInt64(&r.srStats.ReplicaCount),
 	}
@@ -333,6 +336,7 @@ func NewReplicationStats(ctx context.Context, objectAPI ObjectLayer) *Replicatio
 	rs := ReplicationStats{
 		Cache:           make(map[string]*BucketReplicationStats),
 		qCache:          newQueueCache(r),
+		pCache:          newProxyStatsCache(),
 		srStats:         newSRStats(),
 		movingAvgTicker: time.NewTicker(2 * time.Second),
 		wTimer:          time.NewTicker(2 * time.Second),
@@ -371,6 +375,7 @@ func (r *ReplicationStats) calculateBucketReplicationStats(bucket string, bucket
 				Stats: make(map[string]*BucketReplicationStat),
 			},
 			QueueStats: ReplicationQueueStats{},
+			ProxyStats: ProxyMetric{},
 		}
 		return bs
 	}
@@ -430,11 +435,16 @@ func (r *ReplicationStats) calculateBucketReplicationStats(bucket string, bucket
 	for _, bs := range bucketStats {
 		qs.Nodes = append(qs.Nodes, bs.QueueStats.Nodes...)
 	}
-
 	qs.Uptime = UTCNow().Unix() - globalBootTime.Unix()
+
+	var ps ProxyMetric
+	for _, bs := range bucketStats {
+		ps.add(bs.ProxyStats)
+	}
 	bs = BucketStats{
 		ReplicationStats: s,
 		QueueStats:       qs,
+		ProxyStats:       ps,
 	}
 	r.mostRecentStatsMu.Lock()
 	if len(r.mostRecentStats.Stats) == 0 {
@@ -482,3 +492,12 @@ func (r *ReplicationStats) decQ(bucket string, sz int64, isDelMarker bool, opTyp
 	atomic.AddInt64(&r.qCache.srQueueStats.nowBytes, -1*sz)
 	atomic.AddInt64(&r.qCache.srQueueStats.nowCount, -1)
 }
+
+// incProxy increments proxy metrics for proxied calls
+func (r *ReplicationStats) incProxy(bucket string, api replProxyAPI, isErr bool) {
+	r.pCache.inc(bucket, api, isErr)
+}
+
+func (r *ReplicationStats) getProxyStats(bucket string) ProxyMetric {
+	return r.pCache.getBucketStats(bucket)
+}

cmd/bucket-replication-utils.go

@@ -219,7 +219,7 @@ type ReplicateDecision struct {
 	targetsMap map[string]replicateTargetDecision
 }
 
-// ReplicateAny returns true if atleast one target qualifies for replication
+// ReplicateAny returns true if at least one target qualifies for replication
 func (d ReplicateDecision) ReplicateAny() bool {
 	for _, t := range d.targetsMap {
 		if t.Replicate {
@@ -229,7 +229,7 @@ func (d ReplicateDecision) ReplicateAny() bool {
 	return false
 }
 
-// Synchronous returns true if atleast one target qualifies for synchronous replication
+// Synchronous returns true if at least one target qualifies for synchronous replication
 func (d ReplicateDecision) Synchronous() bool {
 	for _, t := range d.targetsMap {
 		if t.Synchronous {
@@ -336,7 +336,7 @@ type ReplicationState struct {
 	DeleteMarker              bool                   // represents DeleteMarker replication state
 	ReplicationTimeStamp      time.Time              // timestamp when last replication activity happened
 	ReplicationStatusInternal string                 // stringified representation of all replication activity
-	// VersionPurgeStatusInternal is internally in the format "arn1=PENDING;arn2=COMMPLETED;"
+	// VersionPurgeStatusInternal is internally in the format "arn1=PENDING;arn2=COMPLETED;"
 	VersionPurgeStatusInternal string                            // stringified representation of all version purge statuses
 	ReplicateDecisionStr       string                            // stringified representation of replication decision for each target
 	Targets                    map[string]replication.StatusType // map of ARN->replication status for ongoing replication activity

cmd/bucket-replication.go

@@ -23,7 +23,6 @@ import (
 	"errors"
 	"fmt"
 	"io"
-	"math"
 	"math/rand"
 	"net/http"
 	"net/url"
@@ -48,6 +47,7 @@ import (
 	"github.com/minio/minio/internal/event"
 	"github.com/minio/minio/internal/hash"
 	xhttp "github.com/minio/minio/internal/http"
+	xioutil "github.com/minio/minio/internal/ioutil"
 	"github.com/minio/minio/internal/logger"
 	"github.com/tinylib/msgp/msgp"
 	"github.com/zeebo/xxh3"
@@ -113,7 +113,7 @@ func validateReplicationDestination(ctx context.Context, bucket string, rCfg *re
 		if arn.Type != madmin.ReplicationService {
 			return sameTarget, toAPIError(ctx, BucketRemoteArnTypeInvalid{Bucket: bucket})
 		}
-		clnt := globalBucketTargetSys.GetRemoteTargetClient(arnStr)
+		clnt := globalBucketTargetSys.GetRemoteTargetClient(bucket, arnStr)
 		if clnt == nil {
 			return sameTarget, toAPIError(ctx, BucketRemoteTargetNotFound{Bucket: bucket})
 		}
@@ -138,7 +138,7 @@ func validateReplicationDestination(ctx context.Context, bucket string, rCfg *re
 			}
 		}
 		// validate replication ARN against target endpoint
-		c := globalBucketTargetSys.GetRemoteTargetClient(arnStr)
+		c := globalBucketTargetSys.GetRemoteTargetClient(bucket, arnStr)
 		if c != nil {
 			if err := checkRemoteEndpoint(ctx, c.EndpointURL()); err != nil {
 				switch err.(type) {
@@ -179,7 +179,7 @@ func checkRemoteEndpoint(ctx context.Context, epURL *url.URL) error {
 	}
 
 	client := &http.Client{
-		Transport: NewHTTPTransport(),
+		Transport: globalRemoteTargetTransport,
 		Timeout:   10 * time.Second,
 	}
 
@@ -280,7 +280,7 @@ func mustReplicate(ctx context.Context, bucket, object string, mopts mustReplica
 	}
 	tgtArns := cfg.FilterTargetArns(opts)
 	for _, tgtArn := range tgtArns {
-		tgt := globalBucketTargetSys.GetRemoteTargetClient(tgtArn)
+		tgt := globalBucketTargetSys.GetRemoteTargetClient(bucket, tgtArn)
 		// the target online status should not be used here while deciding
 		// whether to replicate as the target could be temporarily down
 		opts.TargetArn = tgtArn
@@ -373,15 +373,16 @@ func checkReplicateDelete(ctx context.Context, bucket string, dobj ObjectToDelet
 			if oi.DeleteMarker && (validReplStatus || replicate) {
 				dsc.Set(newReplicateTargetDecision(tgtArn, replicate, sync))
 				continue
-			} else {
-				// can be the case that other cluster is down and duplicate `mc rm --vid`
-				// is issued - this still needs to be replicated back to the other target
+			}
+			// can be the case that other cluster is down and duplicate `mc rm --vid`
+			// is issued - this still needs to be replicated back to the other target
+			if !oi.VersionPurgeStatus.Empty() {
 				replicate = oi.VersionPurgeStatus == Pending || oi.VersionPurgeStatus == Failed
 				dsc.Set(newReplicateTargetDecision(tgtArn, replicate, sync))
-				continue
 			}
+			continue
 		}
-		tgt := globalBucketTargetSys.GetRemoteTargetClient(tgtArn)
+		tgt := globalBucketTargetSys.GetRemoteTargetClient(bucket, tgtArn)
 		// the target online status should not be used here while deciding
 		// whether to replicate deletes as the target could be temporarily down
 		tgtDsc := newReplicateTargetDecision(tgtArn, false, false)
@@ -493,9 +494,9 @@ func replicateDelete(ctx context.Context, dobj DeletedObjectReplicationInfo, obj
 		if dobj.TargetArn != "" && dobj.TargetArn != tgtEntry.Arn {
 			continue
 		}
-		tgtClnt := globalBucketTargetSys.GetRemoteTargetClient(tgtEntry.Arn)
+		tgtClnt := globalBucketTargetSys.GetRemoteTargetClient(bucket, tgtEntry.Arn)
 		if tgtClnt == nil {
-			// Skip stale targets if any and log them to be missing atleast once.
+			// Skip stale targets if any and log them to be missing at least once.
 			logger.LogOnceIf(ctx, fmt.Errorf("failed to get target for bucket:%s arn:%s", bucket, tgtEntry.Arn), tgtEntry.Arn)
 			sendEvent(eventArgs{
 				EventName:  event.ObjectReplicationNotTracked,
@@ -637,19 +638,23 @@ func replicateDeleteToTarget(ctx context.Context, dobj DeletedObjectReplicationI
 				IsReplicationReadyForDeleteMarker: true,
 			},
 		})
+		serr := ErrorRespToObjectError(err, dobj.Bucket, dobj.ObjectName, dobj.VersionID)
 		switch {
-		case isErrMethodNotAllowed(ErrorRespToObjectError(err, dobj.Bucket, dobj.ObjectName)):
+		case isErrMethodNotAllowed(serr):
 			// delete marker already replicated
 			if dobj.VersionID == "" && rinfo.VersionPurgeStatus.Empty() {
 				rinfo.ReplicationStatus = replication.Completed
 				return
 			}
-		case isErrObjectNotFound(ErrorRespToObjectError(err, dobj.Bucket, dobj.ObjectName)):
+		case isErrObjectNotFound(serr), isErrVersionNotFound(serr):
 			// version being purged is already not found on target.
 			if !rinfo.VersionPurgeStatus.Empty() {
 				rinfo.VersionPurgeStatus = Complete
 				return
 			}
+		case isErrReadQuorum(serr), isErrWriteQuorum(serr):
+			// destination has some quorum issues, perform removeObject() anyways
+			// to complete the operation.
 		default:
 			if err != nil && minio.IsNetworkOrHostDown(err, true) && !globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 				globalBucketTargetSys.markOffline(tgt.EndpointURL())
@@ -679,7 +684,7 @@ func replicateDeleteToTarget(ctx context.Context, dobj DeletedObjectReplicationI
 		} else {
 			rinfo.VersionPurgeStatus = Failed
 		}
-		logger.LogIf(ctx, fmt.Errorf("Unable to replicate delete marker to %s/%s(%s): %s", tgt.Bucket, dobj.ObjectName, versionID, rmErr))
+		logger.LogIf(ctx, fmt.Errorf("unable to replicate delete marker to %s: %s/%s(%s): %w", tgt.EndpointURL(), tgt.Bucket, dobj.ObjectName, versionID, rmErr))
 		if rmErr != nil && minio.IsNetworkOrHostDown(rmErr, true) && !globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 			globalBucketTargetSys.markOffline(tgt.EndpointURL())
 		}
@@ -775,6 +780,7 @@ func putReplicationOpts(ctx context.Context, sc string, objInfo ObjectInfo) (put
 		UserMetadata:    meta,
 		ContentType:     objInfo.ContentType,
 		ContentEncoding: objInfo.ContentEncoding,
+		Expires:         objInfo.Expires,
 		StorageClass:    sc,
 		Internal: minio.AdvancedPutOptions{
 			SourceVersionID:    objInfo.VersionID,
@@ -1021,7 +1027,7 @@ func replicateObject(ctx context.Context, ri ReplicateObjectInfo, objectAPI Obje
 	var wg sync.WaitGroup
 	var mu sync.Mutex
 	for _, tgtArn := range tgtArns {
-		tgt := globalBucketTargetSys.GetRemoteTargetClient(tgtArn)
+		tgt := globalBucketTargetSys.GetRemoteTargetClient(bucket, tgtArn)
 		if tgt == nil {
 			logger.LogOnceIf(ctx, fmt.Errorf("failed to get target for bucket:%s arn:%s", bucket, tgtArn), tgtArn)
 			sendEvent(eventArgs{
@@ -1199,7 +1205,7 @@ func (ri ReplicateObjectInfo) replicateObject(ctx context.Context, objectAPI Obj
 	}
 
 	if tgt.Bucket == "" {
-		logger.LogIf(ctx, fmt.Errorf("unable to replicate object %s(%s), bucket is empty", objInfo.Name, objInfo.VersionID))
+		logger.LogIf(ctx, fmt.Errorf("unable to replicate object %s(%s), bucket is empty for target %s", objInfo.Name, objInfo.VersionID, tgt.EndpointURL()))
 		sendEvent(eventArgs{
 			EventName:  event.ObjectReplicationNotTracked,
 			BucketName: bucket,
@@ -1225,7 +1231,7 @@ func (ri ReplicateObjectInfo) replicateObject(ctx context.Context, objectAPI Obj
 
 	putOpts, err := putReplicationOpts(ctx, tgt.StorageClass, objInfo)
 	if err != nil {
-		logger.LogIf(ctx, fmt.Errorf("failed to get target for replication bucket:%s err:%w", bucket, err))
+		logger.LogIf(ctx, fmt.Errorf("failure setting options for replication bucket:%s err:%w", bucket, err))
 		sendEvent(eventArgs{
 			EventName:  event.ObjectReplicationNotTracked,
 			BucketName: bucket,
@@ -1260,14 +1266,14 @@ func (ri ReplicateObjectInfo) replicateObject(ctx context.Context, objectAPI Obj
 			r, objInfo, putOpts); rinfo.Err != nil {
 			if minio.ToErrorResponse(rinfo.Err).Code != "PreconditionFailed" {
 				rinfo.ReplicationStatus = replication.Failed
-				logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err))
+				logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s (target: %s)", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err, tgt.EndpointURL()))
 			}
 		}
 	} else {
 		if _, rinfo.Err = c.PutObject(ctx, tgt.Bucket, object, r, size, "", "", putOpts); rinfo.Err != nil {
 			if minio.ToErrorResponse(rinfo.Err).Code != "PreconditionFailed" {
 				rinfo.ReplicationStatus = replication.Failed
-				logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err))
+				logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s (target: %s)", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err, tgt.EndpointURL()))
 			}
 		}
 	}
@@ -1331,7 +1337,7 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 				UserAgent:  "Internal: [Replication]",
 				Host:       globalLocalNodeName,
 			})
-			logger.LogIf(ctx, fmt.Errorf("unable to update replicate metadata for %s/%s(%s): %w", bucket, object, objInfo.VersionID, err))
+			logger.LogIf(ctx, fmt.Errorf("unable to replicate to target %s for %s/%s(%s): %w", tgt.EndpointURL(), bucket, object, objInfo.VersionID, err))
 		}
 		return
 	}
@@ -1362,7 +1368,7 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 	}
 
 	if tgt.Bucket == "" {
-		logger.LogIf(ctx, fmt.Errorf("unable to replicate object %s(%s), bucket is empty", objInfo.Name, objInfo.VersionID))
+		logger.LogIf(ctx, fmt.Errorf("unable to replicate object %s(%s) to %s, target bucket is missing", objInfo.Name, objInfo.VersionID, tgt.EndpointURL()))
 		sendEvent(eventArgs{
 			EventName:  event.ObjectReplicationNotTracked,
 			BucketName: bucket,
@@ -1380,7 +1386,6 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 		rinfo.Duration = time.Since(startTime)
 	}()
 
-	rAction = replicateAll
 	oi, cerr := tgt.StatObject(ctx, tgt.Bucket, object, minio.StatObjectOptions{
 		VersionID: objInfo.VersionID,
 		Internal: minio.AdvancedGetOptions{
@@ -1393,7 +1398,7 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 		if rAction == replicateNone {
 			if ri.OpType == replication.ExistingObjectReplicationType &&
 				objInfo.ModTime.Unix() > oi.LastModified.Unix() && objInfo.VersionID == nullVersionID {
-				logger.LogIf(ctx, fmt.Errorf("unable to replicate %s/%s (null). Newer version exists on target", bucket, object))
+				logger.LogIf(ctx, fmt.Errorf("unable to replicate %s/%s (null). Newer version exists on target %s", bucket, object, tgt.EndpointURL()))
 				sendEvent(eventArgs{
 					EventName:  event.ObjectReplicationNotTracked,
 					BucketName: bucket,
@@ -1417,16 +1422,19 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 			}
 			return
 		}
-	}
-	// if target returns error other than NoSuchKey, defer replication attempt
-	if cerr != nil {
+	} else {
+		// if target returns error other than NoSuchKey, defer replication attempt
 		if minio.IsNetworkOrHostDown(cerr, true) && !globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 			globalBucketTargetSys.markOffline(tgt.EndpointURL())
 		}
 
-		errResp := minio.ToErrorResponse(cerr)
-		switch errResp.Code {
-		case "NoSuchKey", "NoSuchVersion", "SlowDownRead":
+		serr := ErrorRespToObjectError(cerr, bucket, object, objInfo.VersionID)
+		switch {
+		case isErrMethodNotAllowed(serr):
+			rAction = replicateAll
+		case isErrObjectNotFound(serr), isErrVersionNotFound(serr):
+			rAction = replicateAll
+		case isErrReadQuorum(serr), isErrWriteQuorum(serr):
 			rAction = replicateAll
 		default:
 			rinfo.Err = cerr
@@ -1460,15 +1468,33 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 				ReplicationRequest: true, // always set this to distinguish between `mc mirror` replication and serverside
 			},
 		}
+		if tagTmStr, ok := objInfo.UserDefined[ReservedMetadataPrefixLower+TaggingTimestamp]; ok {
+			ondiskTimestamp, err := time.Parse(time.RFC3339, tagTmStr)
+			if err == nil {
+				dstOpts.Internal.TaggingTimestamp = ondiskTimestamp
+			}
+		}
+		if retTmStr, ok := objInfo.UserDefined[ReservedMetadataPrefixLower+ObjectLockRetentionTimestamp]; ok {
+			ondiskTimestamp, err := time.Parse(time.RFC3339, retTmStr)
+			if err == nil {
+				dstOpts.Internal.RetentionTimestamp = ondiskTimestamp
+			}
+		}
+		if lholdTmStr, ok := objInfo.UserDefined[ReservedMetadataPrefixLower+ObjectLockLegalHoldTimestamp]; ok {
+			ondiskTimestamp, err := time.Parse(time.RFC3339, lholdTmStr)
+			if err == nil {
+				dstOpts.Internal.LegalholdTimestamp = ondiskTimestamp
+			}
+		}
 		if _, rinfo.Err = c.CopyObject(ctx, tgt.Bucket, object, tgt.Bucket, object, getCopyObjMetadata(objInfo, tgt.StorageClass), srcOpts, dstOpts); rinfo.Err != nil {
 			rinfo.ReplicationStatus = replication.Failed
-			logger.LogIf(ctx, fmt.Errorf("unable to replicate metadata for object %s/%s(%s): %s", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err))
+			logger.LogIf(ctx, fmt.Errorf("unable to replicate metadata for object %s/%s(%s) to target %s: %w", bucket, objInfo.Name, objInfo.VersionID, tgt.EndpointURL(), rinfo.Err))
 		}
 	} else {
 		var putOpts minio.PutObjectOptions
 		putOpts, err = putReplicationOpts(ctx, tgt.StorageClass, objInfo)
 		if err != nil {
-			logger.LogIf(ctx, fmt.Errorf("failed to get target for replication bucket:%s err:%w", bucket, err))
+			logger.LogIf(ctx, fmt.Errorf("failed to set replicate options for object %s/%s(%s) (target %s) err:%w", bucket, objInfo.Name, objInfo.VersionID, tgt.EndpointURL(), err))
 			sendEvent(eventArgs{
 				EventName:  event.ObjectReplicationNotTracked,
 				BucketName: bucket,
@@ -1502,7 +1528,7 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 				r, objInfo, putOpts); rinfo.Err != nil {
 				if minio.ToErrorResponse(rinfo.Err).Code != "PreconditionFailed" {
 					rinfo.ReplicationStatus = replication.Failed
-					logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err))
+					logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s) to target %s: %w", bucket, objInfo.Name, objInfo.VersionID, tgt.EndpointURL(), rinfo.Err))
 				} else {
 					rinfo.ReplicationStatus = replication.Completed
 				}
@@ -1511,7 +1537,7 @@ func (ri ReplicateObjectInfo) replicateAll(ctx context.Context, objectAPI Object
 			if _, rinfo.Err = c.PutObject(ctx, tgt.Bucket, object, r, size, "", "", putOpts); rinfo.Err != nil {
 				if minio.ToErrorResponse(rinfo.Err).Code != "PreconditionFailed" {
 					rinfo.ReplicationStatus = replication.Failed
-					logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s): %s", bucket, objInfo.Name, objInfo.VersionID, rinfo.Err))
+					logger.LogIf(ctx, fmt.Errorf("unable to replicate for object %s/%s(%s) to target %s: %w", bucket, objInfo.Name, objInfo.VersionID, tgt.EndpointURL(), rinfo.Err))
 				} else {
 					rinfo.ReplicationStatus = replication.Completed
 				}
@@ -1689,12 +1715,13 @@ type ReplicationPool struct {
 	activeWorkers    int32
 	activeMRFWorkers int32
 
-	objLayer ObjectLayer
-	ctx      context.Context
-	priority string
-	mu       sync.RWMutex
-	mrfMU    sync.Mutex
-	resyncer *replicationResyncer
+	objLayer   ObjectLayer
+	ctx        context.Context
+	priority   string
+	maxWorkers int
+	mu         sync.RWMutex
+	mrfMU      sync.Mutex
+	resyncer   *replicationResyncer
 
 	// workers:
 	workers    []chan ReplicationWorkerOperation
@@ -1740,9 +1767,13 @@ const (
 func NewReplicationPool(ctx context.Context, o ObjectLayer, opts replicationPoolOpts) *ReplicationPool {
 	var workers, failedWorkers int
 	priority := "auto"
+	maxWorkers := WorkerMaxLimit
 	if opts.Priority != "" {
 		priority = opts.Priority
 	}
+	if opts.MaxWorkers > 0 {
+		maxWorkers = opts.MaxWorkers
+	}
 	switch priority {
 	case "fast":
 		workers = WorkerMaxLimit
@@ -1754,7 +1785,13 @@ func NewReplicationPool(ctx context.Context, o ObjectLayer, opts replicationPool
 		workers = WorkerAutoDefault
 		failedWorkers = MRFWorkerAutoDefault
 	}
+	if maxWorkers > 0 && workers > maxWorkers {
+		workers = maxWorkers
+	}
 
+	if maxWorkers > 0 && failedWorkers > maxWorkers {
+		failedWorkers = maxWorkers
+	}
 	pool := &ReplicationPool{
 		workers:         make([]chan ReplicationWorkerOperation, 0, workers),
 		lrgworkers:      make([]chan ReplicationWorkerOperation, 0, LargeWorkerCount),
@@ -1766,6 +1803,7 @@ func NewReplicationPool(ctx context.Context, o ObjectLayer, opts replicationPool
 		ctx:             ctx,
 		objLayer:        o,
 		priority:        priority,
+		maxWorkers:      maxWorkers,
 	}
 
 	pool.AddLargeWorkers()
@@ -1857,7 +1895,7 @@ func (p *ReplicationPool) AddLargeWorkers() {
 	go func() {
 		<-p.ctx.Done()
 		for i := 0; i < LargeWorkerCount; i++ {
-			close(p.lrgworkers[i])
+			xioutil.SafeClose(p.lrgworkers[i])
 		}
 	}()
 }
@@ -1916,12 +1954,12 @@ func (p *ReplicationPool) ResizeWorkers(n, checkOld int) {
 	for len(p.workers) > n {
 		worker := p.workers[len(p.workers)-1]
 		p.workers = p.workers[:len(p.workers)-1]
-		close(worker)
+		xioutil.SafeClose(worker)
 	}
 }
 
 // ResizeWorkerPriority sets replication failed workers pool size
-func (p *ReplicationPool) ResizeWorkerPriority(pri string) {
+func (p *ReplicationPool) ResizeWorkerPriority(pri string, maxWorkers int) {
 	var workers, mrfWorkers int
 	p.mu.Lock()
 	switch pri {
@@ -1935,13 +1973,21 @@ func (p *ReplicationPool) ResizeWorkerPriority(pri string) {
 		workers = WorkerAutoDefault
 		mrfWorkers = MRFWorkerAutoDefault
 		if len(p.workers) < WorkerAutoDefault {
-			workers = int(math.Min(float64(len(p.workers)+1), WorkerAutoDefault))
+			workers = min(len(p.workers)+1, WorkerAutoDefault)
 		}
 		if p.mrfWorkerSize < MRFWorkerAutoDefault {
-			mrfWorkers = int(math.Min(float64(p.mrfWorkerSize+1), MRFWorkerAutoDefault))
+			mrfWorkers = min(p.mrfWorkerSize+1, MRFWorkerAutoDefault)
 		}
 	}
+	if maxWorkers > 0 && workers > maxWorkers {
+		workers = maxWorkers
+	}
+
+	if maxWorkers > 0 && mrfWorkers > maxWorkers {
+		mrfWorkers = maxWorkers
+	}
 	p.priority = pri
+	p.maxWorkers = maxWorkers
 	p.mu.Unlock()
 	p.ResizeWorkers(workers, 0)
 	p.ResizeFailedWorkers(mrfWorkers)
@@ -2015,6 +2061,7 @@ func (p *ReplicationPool) queueReplicaTask(ri ReplicateObjectInfo) {
 		globalReplicationPool.queueMRFSave(ri.ToMRFEntry())
 		p.mu.RLock()
 		prio := p.priority
+		maxWorkers := p.maxWorkers
 		p.mu.RUnlock()
 		switch prio {
 		case "fast":
@@ -2022,16 +2069,18 @@ func (p *ReplicationPool) queueReplicaTask(ri ReplicateObjectInfo) {
 		case "slow":
 			logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Unable to keep up with incoming traffic - we recommend increasing replication priority with `mc admin config set api replication_priority=auto`"), string(replicationSubsystem))
 		default:
-			if p.ActiveWorkers() < WorkerMaxLimit {
+			maxWorkers = min(maxWorkers, WorkerMaxLimit)
+			if p.ActiveWorkers() < maxWorkers {
 				p.mu.RLock()
-				workers := int(math.Min(float64(len(p.workers)+1), WorkerMaxLimit))
+				workers := min(len(p.workers)+1, maxWorkers)
 				existing := len(p.workers)
 				p.mu.RUnlock()
 				p.ResizeWorkers(workers, existing)
 			}
-			if p.ActiveMRFWorkers() < MRFWorkerMaxLimit {
+			maxMRFWorkers := min(maxWorkers, MRFWorkerMaxLimit)
+			if p.ActiveMRFWorkers() < maxMRFWorkers {
 				p.mu.RLock()
-				workers := int(math.Min(float64(p.mrfWorkerSize+1), MRFWorkerMaxLimit))
+				workers := min(p.mrfWorkerSize+1, maxMRFWorkers)
 				p.mu.RUnlock()
 				p.ResizeFailedWorkers(workers)
 			}
@@ -2069,6 +2118,7 @@ func (p *ReplicationPool) queueReplicaDeleteTask(doi DeletedObjectReplicationInf
 		globalReplicationPool.queueMRFSave(doi.ToMRFEntry())
 		p.mu.RLock()
 		prio := p.priority
+		maxWorkers := p.maxWorkers
 		p.mu.RUnlock()
 		switch prio {
 		case "fast":
@@ -2076,9 +2126,10 @@ func (p *ReplicationPool) queueReplicaDeleteTask(doi DeletedObjectReplicationInf
 		case "slow":
 			logger.LogOnceIf(GlobalContext, fmt.Errorf("WARNING: Unable to keep up with incoming deletes - we recommend increasing replication priority with `mc admin config set api replication_priority=auto`"), string(replicationSubsystem))
 		default:
-			if p.ActiveWorkers() < WorkerMaxLimit {
+			maxWorkers = min(maxWorkers, WorkerMaxLimit)
+			if p.ActiveWorkers() < maxWorkers {
 				p.mu.RLock()
-				workers := int(math.Min(float64(len(p.workers)+1), WorkerMaxLimit))
+				workers := min(len(p.workers)+1, maxWorkers)
 				existing := len(p.workers)
 				p.mu.RUnlock()
 				p.ResizeWorkers(workers, existing)
@@ -2088,13 +2139,12 @@ func (p *ReplicationPool) queueReplicaDeleteTask(doi DeletedObjectReplicationInf
 }
 
 type replicationPoolOpts struct {
-	Priority string
+	Priority   string
+	MaxWorkers int
 }
 
 func initBackgroundReplication(ctx context.Context, objectAPI ObjectLayer) {
-	globalReplicationPool = NewReplicationPool(ctx, objectAPI, replicationPoolOpts{
-		Priority: globalAPIConfig.getReplicationPriority(),
-	})
+	globalReplicationPool = NewReplicationPool(ctx, objectAPI, globalAPIConfig.getReplicationOpts())
 	globalReplicationStats = NewReplicationStats(ctx, objectAPI)
 	go globalReplicationStats.trackEWMA()
 }
@@ -2185,8 +2235,9 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRa
 	if opts.ProxyRequest || (opts.ProxyHeaderSet && !opts.ProxyRequest) { // true only when site B sets MinIOSourceProxyRequest header
 		return nil, oi, proxy
 	}
+	var perr error
 	for _, t := range proxyTargets.Targets {
-		tgt = globalBucketTargetSys.GetRemoteTargetClient(t.Arn)
+		tgt = globalBucketTargetSys.GetRemoteTargetClient(bucket, t.Arn)
 		if tgt == nil || globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
 			continue
 		}
@@ -2206,7 +2257,7 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRa
 		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))
+				logger.LogIf(ctx, fmt.Errorf("invalid range header for %s/%s(%s) - %w", bucket, object, opts.VersionID, err))
 				continue
 			}
 			gopts.Set(xhttp.Range, h)
@@ -2214,6 +2265,7 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRa
 
 		objInfo, err := tgt.StatObject(ctx, t.TargetBucket, object, gopts)
 		if err != nil {
+			perr = err
 			if isErrInvalidRange(ErrorRespToObjectError(err, bucket, object)) {
 				return nil, oi, proxyResult{Err: err}
 			}
@@ -2250,6 +2302,7 @@ func proxyHeadToRepTarget(ctx context.Context, bucket, object string, rs *HTTPRa
 		}
 		return tgt, oi, proxyResult{Proxy: true}
 	}
+	proxy.Err = perr
 	return nil, oi, proxy
 }
 
@@ -2299,6 +2352,129 @@ func scheduleReplication(ctx context.Context, oi ObjectInfo, o ObjectLayer, dsc
 	}
 }
 
+// proxyTaggingToRepTarget proxies tagging requests to remote targets for
+// active-active replicated setups
+func proxyTaggingToRepTarget(ctx context.Context, bucket, object string, tags *tags.Tags, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (proxy proxyResult) {
+	// this option is set when active-active replication is in place between site A -> B,
+	// and request hits site B that does not have the object yet.
+	if opts.ProxyRequest || (opts.ProxyHeaderSet && !opts.ProxyRequest) { // true only when site B sets MinIOSourceProxyRequest header
+		return proxy
+	}
+	var wg sync.WaitGroup
+	errs := make([]error, len(proxyTargets.Targets))
+	for idx, t := range proxyTargets.Targets {
+		tgt := globalBucketTargetSys.GetRemoteTargetClient(bucket, t.Arn)
+		if tgt == nil || globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
+			continue
+		}
+		// if proxying explicitly disabled on remote target
+		if tgt.disableProxy {
+			continue
+		}
+		idx := idx
+		wg.Add(1)
+		go func(idx int, tgt *TargetClient) {
+			defer wg.Done()
+			var err error
+			if tags != nil {
+				popts := minio.PutObjectTaggingOptions{
+					VersionID: opts.VersionID,
+					Internal: minio.AdvancedObjectTaggingOptions{
+						ReplicationProxyRequest: "true",
+					},
+				}
+				err = tgt.PutObjectTagging(ctx, tgt.Bucket, object, tags, popts)
+			} else {
+				dopts := minio.RemoveObjectTaggingOptions{
+					VersionID: opts.VersionID,
+					Internal: minio.AdvancedObjectTaggingOptions{
+						ReplicationProxyRequest: "true",
+					},
+				}
+				err = tgt.RemoveObjectTagging(ctx, tgt.Bucket, object, dopts)
+			}
+			if err != nil {
+				errs[idx] = err
+			}
+		}(idx, tgt)
+	}
+	wg.Wait()
+
+	var (
+		terr        error
+		taggedCount int
+	)
+	for _, err := range errs {
+		if err == nil {
+			taggedCount++
+			continue
+		}
+		if err != nil {
+			terr = err
+		}
+	}
+	// don't return error if at least one target was tagged successfully
+	if taggedCount == 0 && terr != nil {
+		proxy.Err = terr
+	}
+	return proxy
+}
+
+// proxyGetTaggingToRepTarget proxies get tagging requests to remote targets for
+// active-active replicated setups
+func proxyGetTaggingToRepTarget(ctx context.Context, bucket, object string, opts ObjectOptions, proxyTargets *madmin.BucketTargets) (tgs *tags.Tags, proxy proxyResult) {
+	// this option is set when active-active replication is in place between site A -> B,
+	// and request hits site B that does not have the object yet.
+	if opts.ProxyRequest || (opts.ProxyHeaderSet && !opts.ProxyRequest) { // true only when site B sets MinIOSourceProxyRequest header
+		return nil, proxy
+	}
+	var wg sync.WaitGroup
+	errs := make([]error, len(proxyTargets.Targets))
+	tagSlc := make([]map[string]string, len(proxyTargets.Targets))
+	for idx, t := range proxyTargets.Targets {
+		tgt := globalBucketTargetSys.GetRemoteTargetClient(bucket, t.Arn)
+		if tgt == nil || globalBucketTargetSys.isOffline(tgt.EndpointURL()) {
+			continue
+		}
+		// if proxying explicitly disabled on remote target
+		if tgt.disableProxy {
+			continue
+		}
+		idx := idx
+		wg.Add(1)
+		go func(idx int, tgt *TargetClient) {
+			defer wg.Done()
+			var err error
+			gopts := minio.GetObjectTaggingOptions{
+				VersionID: opts.VersionID,
+				Internal: minio.AdvancedObjectTaggingOptions{
+					ReplicationProxyRequest: "true",
+				},
+			}
+			tgs, err = tgt.GetObjectTagging(ctx, tgt.Bucket, object, gopts)
+			if err != nil {
+				errs[idx] = err
+			} else {
+				tagSlc[idx] = tgs.ToMap()
+			}
+		}(idx, tgt)
+	}
+	wg.Wait()
+	for idx, err := range errs {
+		errCode := minio.ToErrorResponse(err).Code
+		if err != nil && errCode != "NoSuchKey" && errCode != "NoSuchVersion" {
+			return nil, proxyResult{Err: err}
+		}
+		if err == nil {
+			tgs, _ = tags.MapToObjectTags(tagSlc[idx])
+		}
+	}
+	if len(errs) == 1 {
+		proxy.Err = errs[0]
+	}
+	return tgs, proxy
+}
+
 func scheduleReplicationDelete(ctx context.Context, dv DeletedObjectReplicationInfo, o ObjectLayer) {
 	globalReplicationPool.queueReplicaDeleteTask(dv)
 	for arn := range dv.ReplicationState.Targets {
@@ -2449,7 +2625,7 @@ func (s *replicationResyncer) PersistToDisk(ctx context.Context, objectAPI Objec
 				}
 				if updt {
 					if err := saveResyncStatus(ctx, bucket, brs, objectAPI); err != nil {
-						logger.LogIf(ctx, fmt.Errorf("Could not save resync metadata to drive for %s - %w", bucket, err))
+						logger.LogIf(ctx, fmt.Errorf("could not save resync metadata to drive for %s - %w", bucket, err))
 					} else {
 						lastResyncStatusSave[bucket] = brs.LastUpdate
 					}
@@ -2485,7 +2661,7 @@ func newresyncer() *replicationResyncer {
 }
 
 // mark status of replication resync on remote target for the bucket
-func (s *replicationResyncer) markStatus(status ResyncStatusType, opts resyncOpts) {
+func (s *replicationResyncer) markStatus(status ResyncStatusType, opts resyncOpts, objAPI ObjectLayer) {
 	s.Lock()
 	defer s.Unlock()
 
@@ -2496,6 +2672,10 @@ func (s *replicationResyncer) markStatus(status ResyncStatusType, opts resyncOpt
 	m.TargetsMap[opts.arn] = st
 	m.LastUpdate = UTCNow()
 	s.statusMap[opts.bucket] = m
+
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second)
+	defer cancel()
+	saveResyncStatus(ctx, opts.bucket, m, objAPI)
 }
 
 // update replication resync stats for bucket's remote target
@@ -2525,7 +2705,7 @@ func (s *replicationResyncer) resyncBucket(ctx context.Context, objectAPI Object
 
 	resyncStatus := ResyncFailed
 	defer func() {
-		s.markStatus(resyncStatus, opts)
+		s.markStatus(resyncStatus, opts, objectAPI)
 		globalSiteResyncMetrics.incBucket(opts, resyncStatus)
 		s.workerCh <- struct{}{}
 	}()
@@ -2554,19 +2734,19 @@ func (s *replicationResyncer) resyncBucket(ctx context.Context, objectAPI Object
 		logger.LogIf(ctx, fmt.Errorf("replication resync failed for %s - arn specified %s is missing in the replication config", opts.bucket, opts.arn))
 		return
 	}
-	tgt := globalBucketTargetSys.GetRemoteTargetClient(opts.arn)
+	tgt := globalBucketTargetSys.GetRemoteTargetClient(opts.bucket, opts.arn)
 	if tgt == nil {
 		logger.LogIf(ctx, fmt.Errorf("replication resync failed for %s - target could not be created for arn %s", opts.bucket, opts.arn))
 		return
 	}
 	// mark resync status as resync started
 	if !heal {
-		s.markStatus(ResyncStarted, opts)
+		s.markStatus(ResyncStarted, opts, objectAPI)
 	}
 
 	// Walk through all object versions - Walk() is always in ascending order needed to ensure
 	// delete marker replicated to target after object version is first created.
-	if err := objectAPI.Walk(ctx, opts.bucket, "", objInfoCh, ObjectOptions{}); err != nil {
+	if err := objectAPI.Walk(ctx, opts.bucket, "", objInfoCh, WalkOptions{}); err != nil {
 		logger.LogIf(ctx, err)
 		return
 	}
@@ -2581,7 +2761,7 @@ func (s *replicationResyncer) resyncBucket(ctx context.Context, objectAPI Object
 	}
 	workers := make([]chan ReplicateObjectInfo, resyncParallelRoutines)
 	resultCh := make(chan TargetReplicationResyncStatus, 1)
-	defer close(resultCh)
+	defer xioutil.SafeClose(resultCh)
 	go func() {
 		for r := range resultCh {
 			s.incStats(r, opts)
@@ -2632,16 +2812,18 @@ func (s *replicationResyncer) resyncBucket(ctx context.Context, objectAPI Object
 					roi.EventType = ReplicateExisting
 					replicateObject(ctx, roi, objectAPI)
 				}
-				_, err = tgt.StatObject(ctx, tgt.Bucket, roi.Name, minio.StatObjectOptions{
+
+				st := TargetReplicationResyncStatus{
+					Object: roi.Name,
+					Bucket: roi.Bucket,
+				}
+
+				_, err := tgt.StatObject(ctx, tgt.Bucket, roi.Name, minio.StatObjectOptions{
 					VersionID: roi.VersionID,
 					Internal: minio.AdvancedGetOptions{
 						ReplicationProxyRequest: "false",
 					},
 				})
-				st := TargetReplicationResyncStatus{
-					Object: roi.Name,
-					Bucket: roi.Bucket,
-				}
 				if err != nil {
 					if roi.DeleteMarker && isErrMethodNotAllowed(ErrorRespToObjectError(err, opts.bucket, roi.Name)) {
 						st.ReplicatedCount++
@@ -2691,7 +2873,7 @@ func (s *replicationResyncer) resyncBucket(ctx context.Context, objectAPI Object
 		}
 	}
 	for i := 0; i < resyncParallelRoutines; i++ {
-		close(workers[i])
+		xioutil.SafeClose(workers[i])
 	}
 	wg.Wait()
 	resyncStatus = ResyncCompleted
@@ -2821,7 +3003,7 @@ func (p *ReplicationPool) startResyncRoutine(ctx context.Context, buckets []Buck
 		}
 		duration := time.Duration(r.Float64() * float64(time.Minute))
 		if duration < time.Second {
-			// Make sure to sleep atleast a second to avoid high CPU ticks.
+			// Make sure to sleep at least a second to avoid high CPU ticks.
 			duration = time.Second
 		}
 		time.Sleep(duration)
@@ -2937,7 +3119,7 @@ func getReplicationDiff(ctx context.Context, objAPI ObjectLayer, bucket string,
 	}
 
 	objInfoCh := make(chan ObjectInfo, 10)
-	if err := objAPI.Walk(ctx, bucket, opts.Prefix, objInfoCh, ObjectOptions{}); err != nil {
+	if err := objAPI.Walk(ctx, bucket, opts.Prefix, objInfoCh, WalkOptions{}); err != nil {
 		logger.LogIf(ctx, err)
 		return nil, err
 	}
@@ -2947,7 +3129,7 @@ func getReplicationDiff(ctx context.Context, objAPI ObjectLayer, bucket string,
 	}
 	diffCh := make(chan madmin.DiffInfo, 4000)
 	go func() {
-		defer close(diffCh)
+		defer xioutil.SafeClose(diffCh)
 		for obj := range objInfoCh {
 			if contextCanceled(ctx) {
 				// Just consume input...
@@ -3036,6 +3218,9 @@ func queueReplicationHeal(ctx context.Context, bucket string, oi ObjectInfo, rcf
 		return roi
 	}
 
+	if isVeeamSOSAPIObject(oi.Name) {
+		return roi
+	}
 	if rcfg.Config == nil || rcfg.remotes == nil {
 		return roi
 	}
@@ -3140,7 +3325,7 @@ func (p *ReplicationPool) persistMRF() {
 			mTimer.Reset(mrfSaveInterval)
 		case <-p.ctx.Done():
 			p.mrfStopCh <- struct{}{}
-			close(p.mrfSaveCh)
+			xioutil.SafeClose(p.mrfSaveCh)
 			// We try to save if possible, but we don't care beyond that.
 			saveMRFToDisk()
 			return
@@ -3148,10 +3333,11 @@ func (p *ReplicationPool) persistMRF() {
 			if !ok {
 				return
 			}
+			entries[e.versionID] = e
+
 			if len(entries) >= mrfMaxEntries {
 				saveMRFToDisk()
 			}
-			entries[e.versionID] = e
 		}
 	}
 }
@@ -3175,16 +3361,22 @@ func (p *ReplicationPool) queueMRFSave(entry MRFReplicateEntry) {
 		select {
 		case p.mrfSaveCh <- entry:
 		default:
+			atomic.AddUint64(&globalReplicationStats.mrfStats.TotalDroppedCount, 1)
+			atomic.AddUint64(&globalReplicationStats.mrfStats.TotalDroppedBytes, uint64(entry.sz))
 		}
 	}
 }
 
-func (p *ReplicationPool) persistToDrive(ctx context.Context, v MRFReplicateEntries, data []byte) {
+func (p *ReplicationPool) persistToDrive(ctx context.Context, v MRFReplicateEntries) {
 	newReader := func() io.ReadCloser {
 		r, w := io.Pipe()
 		go func() {
+			// Initialize MRF meta header.
+			var data [4]byte
+			binary.LittleEndian.PutUint16(data[0:2], mrfMetaFormat)
+			binary.LittleEndian.PutUint16(data[2:4], mrfMetaVersion)
 			mw := msgp.NewWriter(w)
-			n, err := mw.Write(data)
+			n, err := mw.Write(data[:])
 			if err != nil {
 				w.CloseWithError(err)
 				return
@@ -3200,9 +3392,13 @@ func (p *ReplicationPool) persistToDrive(ctx context.Context, v MRFReplicateEntr
 		return r
 	}
 
-	for _, localDrive := range globalLocalDrives {
+	globalLocalDrivesMu.RLock()
+	localDrives := globalLocalDrives
+	globalLocalDrivesMu.RUnlock()
+
+	for _, localDrive := range localDrives {
 		r := newReader()
-		err := localDrive.CreateFile(ctx, minioMetaBucket, pathJoin(replicationMRFDir, globalLocalNodeNameHex+".bin"), -1, r)
+		err := localDrive.CreateFile(ctx, "", minioMetaBucket, pathJoin(replicationMRFDir, globalLocalNodeNameHex+".bin"), -1, r)
 		r.Close()
 		if err == nil {
 			break
@@ -3222,15 +3418,10 @@ func (p *ReplicationPool) saveMRFEntries(ctx context.Context, entries map[string
 
 	v := MRFReplicateEntries{
 		Entries: entries,
-		Version: mrfMetaVersionV1,
+		Version: mrfMetaVersion,
 	}
-	data := make([]byte, 4, v.Msgsize()+4)
 
-	// Initialize the resync meta header.
-	binary.LittleEndian.PutUint16(data[0:2], mrfMetaFormat)
-	binary.LittleEndian.PutUint16(data[2:4], mrfMetaVersion)
-
-	p.persistToDrive(ctx, v, data)
+	p.persistToDrive(ctx, v)
 }
 
 // load mrf entries from disk
@@ -3241,8 +3432,8 @@ func (p *ReplicationPool) loadMRF() (mrfRec MRFReplicateEntries, err error) {
 		if !p.initialized() {
 			return re, nil
 		}
-		data := make([]byte, 4)
-		n, err := rc.Read(data)
+		var data [4]byte
+		n, err := rc.Read(data[:])
 		if err != nil {
 			return re, err
 		}
@@ -3267,7 +3458,12 @@ func (p *ReplicationPool) loadMRF() (mrfRec MRFReplicateEntries, err error) {
 
 		return re, nil
 	}
-	for _, localDrive := range globalLocalDrives {
+
+	globalLocalDrivesMu.RLock()
+	localDrives := globalLocalDrives
+	globalLocalDrivesMu.RUnlock()
+
+	for _, localDrive := range localDrives {
 		rc, err := localDrive.ReadFileStream(p.ctx, minioMetaBucket, pathJoin(replicationMRFDir, globalLocalNodeNameHex+".bin"), 0, -1)
 		if err != nil {
 			continue
@@ -3356,7 +3552,7 @@ func (p *ReplicationPool) initialized() bool {
 }
 
 // getMRF returns MRF entries for this node.
-func (p *ReplicationPool) getMRF(ctx context.Context, bucket string) (ch chan madmin.ReplicationMRF, err error) {
+func (p *ReplicationPool) getMRF(ctx context.Context, bucket string) (ch <-chan madmin.ReplicationMRF, err error) {
 	mrfRec, err := p.loadMRF()
 	if err != nil {
 		return nil, err
@@ -3364,9 +3560,9 @@ func (p *ReplicationPool) getMRF(ctx context.Context, bucket string) (ch chan ma
 
 	mrfCh := make(chan madmin.ReplicationMRF, 100)
 	go func() {
-		defer close(mrfCh)
+		defer xioutil.SafeClose(mrfCh)
 		for vID, e := range mrfRec.Entries {
-			if e.Bucket != bucket && bucket != "" {
+			if bucket != "" && e.Bucket != bucket {
 				continue
 			}
 			select {

cmd/bucket-stats.go

@@ -154,6 +154,7 @@ type BucketStats struct {
 	Uptime           int64                  `json:"uptime"`
 	ReplicationStats BucketReplicationStats `json:"currStats"`  // current replication stats since cluster startup
 	QueueStats       ReplicationQueueStats  `json:"queueStats"` // replication queue stats
+	ProxyStats       ProxyMetric            `json:"proxyStats"`
 }
 
 // BucketReplicationStats represents inline replication statistics