Merge branch 'main' into dependabot/go_modules/github.com/moby/spdystream-0.5.1

Unified repo interface extension for block data mover

.github/auto-assignees.yml

@@ -13,10 +13,10 @@ reviewers:
       - reasonerjt
       - ywk253100
       - blackpiglet
-      - qiuming-best
       - shubham-pampattiwar
       - Lyndon-Li
       - anshulahuja98
+      - kaovilai
 
     tech-writer:
       - sseago

.github/pull_request_template.md

@@ -9,5 +9,5 @@ Fixes #(issue)
 # Please indicate you've done the following:
 
 - [ ] [Accepted the DCO](https://velero.io/docs/v1.5/code-standards/#dco-sign-off). Commits without the DCO will delay acceptance.
-- [ ] [Created a changelog file](https://velero.io/docs/v1.5/code-standards/#adding-a-changelog) or added `/kind changelog-not-required` as a comment on this pull request.
+- [ ] [Created a changelog file (`make new-changelog`)](https://velero.io/docs/main/code-standards/#adding-a-changelog) or comment `/kind changelog-not-required` on this PR.
 - [ ] Updated the corresponding documentation in `site/content/docs/main`.

.github/workflows/crds-verify-kind.yaml

@@ -1,93 +0,0 @@
-name: "Verify Velero CRDs across k8s versions"
-on:
-  pull_request:
-    # Do not run when the change only includes these directories.
-    paths-ignore:
-      - "site/**"
-      - "design/**"
-
-jobs:
-  # Build the Velero CLI once for all Kubernetes versions, and cache it so the fan-out workers can get it.
-  build-cli:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Set up Go
-        uses: actions/setup-go@v5
-        with:
-          go-version: '1.22'
-        id: go
-      # Look for a CLI that's made for this PR
-      - name: Fetch built CLI
-        id: cache
-        uses: actions/cache@v4
-        env:
-          cache-name: cache-velero-cli
-        with:
-          path: ./_output/bin/linux/amd64/velero
-          # The cache key a combination of the current PR number, and a SHA256 hash of the Velero binary
-          key: velero-${{ github.event.pull_request.number }}-${{ hashFiles('./_output/bin/linux/amd64/velero') }}
-          # This key controls the prefixes that we'll look at in the cache to restore from
-          restore-keys: |
-            velero-${{ github.event.pull_request.number }}-
-
-      - name: Fetch cached go modules
-        uses: actions/cache@v4
-        if: steps.cache.outputs.cache-hit != 'true'
-        with:
-          path: ~/go/pkg/mod
-          key: ${{ runner.os }}-go-${{ hashFiles('**/go.sum') }}
-          restore-keys: |
-            ${{ runner.os }}-go-
-
-      - name: Check out the code
-        uses: actions/checkout@v4
-        if: steps.cache.outputs.cache-hit != 'true'
-
-      # If no binaries were built for this PR, build it now.
-      - name: Build Velero CLI
-        if: steps.cache.outputs.cache-hit != 'true'
-        run: |
-          make local
-
-  # Check the common CLI against all Kubernetes versions
-  crd-check:
-    needs: build-cli
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        # Latest k8s versions. There's no series-based tag, nor is there a latest tag.
-        k8s:
-          - 1.23.17
-          - 1.24.17
-          - 1.25.16
-          - 1.26.13
-          - 1.27.10
-          - 1.28.6
-          - 1.29.1
-    # All steps run in parallel unless otherwise specified.
-    # See https://docs.github.com/en/actions/learn-github-actions/managing-complex-workflows#creating-dependent-jobs
-    steps:
-      - name: Fetch built CLI
-        id: cache
-        uses: actions/cache@v4
-        env:
-          cache-name: cache-velero-cli
-        with:
-          path: ./_output/bin/linux/amd64/velero
-          # The cache key a combination of the current PR number, and a SHA256 hash of the Velero binary
-          key: velero-${{ github.event.pull_request.number }}-${{ hashFiles('./_output/bin/linux/amd64/velero') }}
-          # This key controls the prefixes that we'll look at in the cache to restore from
-          restore-keys: |
-            velero-${{ github.event.pull_request.number }}-
-      - uses: engineerd/setup-kind@v0.5.0
-        with:
-          version: "v0.21.0"
-          image: "kindest/node:v${{ matrix.k8s }}"
-      - name: Install CRDs
-        run: |
-          kubectl cluster-info
-          kubectl get pods -n kube-system
-          kubectl version
-          echo "current-context:" $(kubectl config current-context)
-          echo "environment-kubeconfig:" ${KUBECONFIG}
-          ./_output/bin/linux/amd64/velero install --crds-only --dry-run -oyaml | kubectl apply -f -

.github/workflows/e2e-test-kind.yaml

@@ -6,16 +6,28 @@ on:
     paths-ignore:
       - "site/**"
       - "design/**"
+      - "**/*.md"
 jobs:
+  get-go-version:
+    uses: ./.github/workflows/get-go-version.yaml
+    with:
+      ref: ${{ github.event.pull_request.base.ref }}
+
   # Build the Velero CLI and image once for all Kubernetes versions, and cache it so the fan-out workers can get it.
   build:
     runs-on: ubuntu-latest
+    needs: get-go-version
+    outputs:
+      minio-dockerfile-sha: ${{ steps.minio-version.outputs.dockerfile_sha }}
     steps:
-      - name: Set up Go
-        uses: actions/setup-go@v5
+      - name: Check out the code
+        uses: actions/checkout@v6
+      
+      - name: Set up Go version
+        uses: actions/setup-go@v6
         with:
-          go-version: '1.22'
-        id: go
+          go-version: ${{ needs.get-go-version.outputs.version }}
+
       # Look for a CLI that's made for this PR
       - name: Fetch built CLI
         id: cli-cache
@@ -31,17 +43,6 @@ jobs:
           path: ./velero.tar
           # The cache key a combination of the current PR number and the commit SHA
           key: velero-image-${{ github.event.pull_request.number }}-${{ github.sha }}
-      - name: Fetch cached go modules
-        uses: actions/cache@v4
-        if: steps.cli-cache.outputs.cache-hit != 'true'
-        with:
-          path: ~/go/pkg/mod
-          key: ${{ runner.os }}-go-${{ hashFiles('**/go.sum') }}
-          restore-keys: |
-            ${{ runner.os }}-go-
-      - name: Check out the code
-        uses: actions/checkout@v4
-        if: steps.cli-cache.outputs.cache-hit != 'true' || steps.image-cache.outputs.cache-hit != 'true'
       # If no binaries were built for this PR, build it now.
       - name: Build Velero CLI
         if: steps.cli-cache.outputs.cache-hit != 'true'
@@ -51,47 +52,91 @@ jobs:
       - name: Build Velero Image
         if: steps.image-cache.outputs.cache-hit != 'true'
         run: |
-          IMAGE=velero VERSION=pr-test make container
-          docker save velero:pr-test -o ./velero.tar
+          IMAGE=velero VERSION=pr-test BUILD_OUTPUT_TYPE=docker make container
+          docker save velero:pr-test-linux-amd64 -o ./velero.tar
+      # Check and build MinIO image once for all e2e tests
+      - name: Check Bitnami MinIO Dockerfile version
+        id: minio-version
+        run: |
+          DOCKERFILE_SHA=$(curl -s https://api.github.com/repos/bitnami/containers/commits?path=bitnami/minio/2025/debian-12/Dockerfile\&per_page=1 | jq -r '.[0].sha')
+          echo "dockerfile_sha=${DOCKERFILE_SHA}" >> $GITHUB_OUTPUT
+      - name: Cache MinIO Image
+        uses: actions/cache@v4
+        id: minio-cache
+        with:
+          path: ./minio-image.tar
+          key: minio-bitnami-${{ steps.minio-version.outputs.dockerfile_sha }}
+      - name: Build MinIO Image from Bitnami Dockerfile
+        if: steps.minio-cache.outputs.cache-hit != 'true'
+        run: |
+          echo "Building MinIO image from Bitnami Dockerfile..."
+          git clone --depth 1 https://github.com/bitnami/containers.git /tmp/bitnami-containers
+          cd /tmp/bitnami-containers/bitnami/minio/2025/debian-12
+          docker build -t bitnami/minio:local .
+          docker save bitnami/minio:local > ${{ github.workspace }}/minio-image.tar
+  # Create json of k8s versions to test
+  # from guide: https://stackoverflow.com/a/65094398/4590470
+  setup-test-matrix:
+    runs-on: ubuntu-latest
+    env:
+      GH_TOKEN: ${{ github.token }}
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+    steps:
+      - name: Set k8s versions
+        id: set-matrix
+        # everything excluding older tags. limits needs to be high enough to cover all latest versions
+        # and test labels
+        # grep -E "v[1-9]\.(2[5-9]|[3-9][0-9])" filters for v1.25 to v9.99
+        # and removes older patches of the same minor version
+        # awk -F. '{if(!a[$1"."$2]++)print $1"."$2"."$NF}'
+        run: |
+          echo "matrix={\
+            \"k8s\":$(wget -q -O - "https://hub.docker.com/v2/namespaces/kindest/repositories/node/tags?page_size=50" | grep -o '"name": *"[^"]*' | grep -o '[^"]*$' | grep -v -E "alpha|beta" | grep -E "v[1-9]\.(2[5-9]|[3-9][0-9])" | awk -F. '{if(!a[$1"."$2]++)print $1"."$2"."$NF}' | sort -r | sed s/v//g | jq -R -c -s 'split("\n")[:-1]'),\
+            \"labels\":[\
+              \"Basic && (ClusterResource || NodePort || StorageClass)\", \
+              \"ResourceFiltering && !Restic\", \
+              \"ResourceModifier || (Backups && BackupsSync) || PrivilegesMgmt || OrderedResources\", \
+              \"(NamespaceMapping && Single && Restic) || (NamespaceMapping && Multiple && Restic)\"\
+            ]}" >> $GITHUB_OUTPUT
+
   # Run E2E test against all Kubernetes versions on kind
   run-e2e-test:
-    needs: build
+    needs:
+      - build
+      - setup-test-matrix
+      - get-go-version
     runs-on: ubuntu-latest
     strategy:
-      matrix:
-        k8s:
-          - 1.23.17
-          - 1.24.17
-          - 1.25.16
-          - 1.26.13
-          - 1.27.10
-          - 1.28.6
-          - 1.29.1
-        focus:
-          # tests to focus on, use `|` to concatenate multiple regexes to run on the same job
-          # ordered according to e2e_suite_test.go order
-          - Basic\]\[ClusterResource
-          - ResourceFiltering
-          - ResourceModifier|Backups|PrivilegesMgmt\]\[SSR
-          - Schedule\]\[OrderedResources
-          - NamespaceMapping\]\[Single\]\[Restic|NamespaceMapping\]\[Multiple\]\[Restic
-          - Basic\]\[Nodeport
-          - Basic\]\[StorageClass
+      matrix: ${{fromJson(needs.setup-test-matrix.outputs.matrix)}}
       fail-fast: false
     steps:
-      - name: Set up Go
-        uses: actions/setup-go@v5
-        with:
-          go-version: '1.22'
-        id: go
       - name: Check out the code
-        uses: actions/checkout@v4
-      - name: Install MinIO
-        run:
-          docker run -d --rm -p 9000:9000 -e "MINIO_ACCESS_KEY=minio" -e "MINIO_SECRET_KEY=minio123" -e "MINIO_DEFAULT_BUCKETS=bucket,additional-bucket" bitnami/minio:2021.6.17-debian-10-r7
-      - uses: engineerd/setup-kind@v0.5.0
+        uses: actions/checkout@v6
+
+      - name: Set up Go version
+        uses: actions/setup-go@v6
         with:
-          version: "v0.21.0"
+          go-version: ${{ needs.get-go-version.outputs.version }}
+
+      # Fetch the pre-built MinIO image from the build job
+      - name: Fetch built MinIO Image
+        uses: actions/cache@v4
+        id: minio-cache
+        with:
+          path: ./minio-image.tar
+          key: minio-bitnami-${{ needs.build.outputs.minio-dockerfile-sha }}
+      - name: Load MinIO Image
+        run: |
+          echo "Loading MinIO image..."
+          docker load < ./minio-image.tar
+      - name: Install MinIO
+        run: |
+          docker run -d --rm -p 9000:9000 -e "MINIO_ROOT_USER=minio" -e "MINIO_ROOT_PASSWORD=minio123" -e "MINIO_DEFAULT_BUCKETS=bucket,additional-bucket" bitnami/minio:local
+      - uses: engineerd/setup-kind@v0.6.2
+        with:
+          skipClusterLogsExport: true
+          version: "v0.27.0"
           image: "kindest/node:v${{ matrix.k8s }}"
       - name: Fetch built CLI
         id: cli-cache
@@ -108,14 +153,6 @@ jobs:
       - name: Load Velero Image
         run:
           kind load image-archive velero.tar
-      # always try to fetch the cached go modules as the e2e test needs it either
-      - name: Fetch cached go modules
-        uses: actions/cache@v4
-        with:
-          path: ~/go/pkg/mod
-          key: ${{ runner.os }}-go-${{ hashFiles('**/go.sum') }}
-          restore-keys: |
-            ${{ runner.os }}-go-
       - name: Run E2E test
         run: |
           cat << EOF > /tmp/credential
@@ -128,18 +165,27 @@ jobs:
           curl -LO https://dl.k8s.io/release/v${{ matrix.k8s }}/bin/linux/amd64/kubectl
           sudo install -o root -g root -m 0755 kubectl /usr/local/bin/kubectl
 
-          GOPATH=~/go CLOUD_PROVIDER=kind \
-              OBJECT_STORE_PROVIDER=aws BSL_CONFIG=region=minio,s3ForcePathStyle="true",s3Url=http://$(hostname -i):9000 \
-              CREDS_FILE=/tmp/credential BSL_BUCKET=bucket \
-              ADDITIONAL_OBJECT_STORE_PROVIDER=aws ADDITIONAL_BSL_CONFIG=region=minio,s3ForcePathStyle="true",s3Url=http://$(hostname -i):9000 \
-              ADDITIONAL_CREDS_FILE=/tmp/credential ADDITIONAL_BSL_BUCKET=additional-bucket \
-              GINKGO_FOCUS='${{ matrix.focus }}' VELERO_IMAGE=velero:pr-test \
-              GINKGO_SKIP='SKIP_KIND|pv-backup|Restic|Snapshot|LongTime' \
+          git clone https://github.com/vmware-tanzu-experiments/distributed-data-generator.git -b main /tmp/kibishii
+
+          GOPATH=~/go \
+              CLOUD_PROVIDER=kind \
+              OBJECT_STORE_PROVIDER=aws \
+              BSL_CONFIG=region=minio,s3ForcePathStyle="true",s3Url=http://$(hostname -i):9000 \
+              CREDS_FILE=/tmp/credential \
+              BSL_BUCKET=bucket \
+              ADDITIONAL_OBJECT_STORE_PROVIDER=aws \
+              ADDITIONAL_BSL_CONFIG=region=minio,s3ForcePathStyle="true",s3Url=http://$(hostname -i):9000 \
+              ADDITIONAL_CREDS_FILE=/tmp/credential \
+              ADDITIONAL_BSL_BUCKET=additional-bucket \
+              VELERO_IMAGE=velero:pr-test-linux-amd64 \
+              PLUGINS=velero/velero-plugin-for-aws:latest \
+              GINKGO_LABELS="${{ matrix.labels }}" \
+              KIBISHII_DIRECTORY=/tmp/kibishii/kubernetes/yaml/ \
               make -C test/ run-e2e
         timeout-minutes: 30
       - name: Upload debug bundle
         if: ${{ failure() }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@v5
         with:
-          name: DebugBundle
-          path: /home/runner/work/velero/velero/test/e2e/debug-bundle*
+          name: DebugBundle-k8s-${{ matrix.k8s }}-job-${{ strategy.job-index }}
+          path: /home/runner/work/velero/velero/test/e2e/debug-bundle*

.github/workflows/get-go-version.yaml

@@ -0,0 +1,33 @@
+on:
+  workflow_call:
+    inputs:
+      ref:
+        description: "The target branch's ref"
+        required: true
+        type: string
+    outputs:
+      version: 
+        description: "The expected Go version"
+        value: ${{ jobs.extract.outputs.version }}
+
+jobs:
+  extract:
+      runs-on: ubuntu-latest
+      outputs:
+        version: ${{ steps.pick-version.outputs.version }}
+      steps:
+        - name: Check out the code
+          uses: actions/checkout@v6
+
+        - id: pick-version
+          run: |
+            if [ "${{ inputs.ref }}" == "main" ]; then
+              version=$(grep '^go ' go.mod | awk '{print $2}' | cut -d. -f1-2)
+            else
+              goDirectiveVersion=$(grep '^go ' go.mod | awk '{print $2}')
+              toolChainVersion=$(grep '^toolchain ' go.mod | awk '{print $2}')
+              version=$(printf "%s\n%s\n" "$goDirectiveVersion" "$toolChainVersion" | sort -V | tail -n1)
+            fi
+
+            echo "version=$version"
+            echo "version=$version" >> $GITHUB_OUTPUT

.github/workflows/nightly-trivy-scan.yml

@@ -13,16 +13,16 @@ jobs:
         # maintain the versions of Velero those need security scan
         versions: [main]
         # list of images that need scan
-        images: [velero, velero-restore-helper]
+        images: [velero, velero-plugin-for-aws, velero-plugin-for-gcp, velero-plugin-for-microsoft-azure]
     permissions:
       security-events: write  # for github/codeql-action/upload-sarif to upload SARIF results
 
     steps:
       - name: Checkout code
-        uses: actions/checkout@v4
+        uses: actions/checkout@v6
 
       - name: Run Trivy vulnerability scanner
-        uses: aquasecurity/trivy-action@master
+        uses: aquasecurity/trivy-action@57a97c7e7821a5776cebc9bb87c984fa69cba8f1
         with:
           image-ref: 'docker.io/velero/${{ matrix.images }}:${{ matrix.versions }}'
           severity: 'CRITICAL,HIGH,MEDIUM'

.github/workflows/pr-changelog-check.yml

@@ -12,7 +12,7 @@ jobs:
     steps:
 
     - name: Check out the code
-      uses: actions/checkout@v4
+      uses: actions/checkout@v6
 
     - name: Changelog check
       if: ${{ !(contains(github.event.pull_request.labels.*.name, 'kind/changelog-not-required') || contains(github.event.pull_request.labels.*.name, 'Design') || contains(github.event.pull_request.labels.*.name, 'Website') || contains(github.event.pull_request.labels.*.name, 'Documentation'))}}

.github/workflows/pr-ci-check.yml

@@ -1,30 +1,30 @@
 name: Pull Request CI Check
 on: [pull_request]
 jobs:
+  get-go-version:
+    uses: ./.github/workflows/get-go-version.yaml
+    with:
+      ref: ${{ github.event.pull_request.base.ref }}
+
   build:
     name: Run CI
+    needs: get-go-version
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
     steps:
-      - name: Set up Go
-        uses: actions/setup-go@v5
-        with:
-          go-version: '1.22'
-        id: go
       - name: Check out the code
-        uses: actions/checkout@v4
-      - name: Fetch cached go modules
-        uses: actions/cache@v4
+        uses: actions/checkout@v6
+
+      - name: Set up Go version
+        uses: actions/setup-go@v6
         with:
-          path: ~/go/pkg/mod
-          key: ${{ runner.os }}-go-${{ hashFiles('**/go.sum') }}
-          restore-keys: |
-            ${{ runner.os }}-go-
+          go-version: ${{ needs.get-go-version.outputs.version }}      
+
       - name: Make ci
         run: make ci
       - name: Upload test coverage
-        uses: codecov/codecov-action@v4
+        uses: codecov/codecov-action@v5
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           files: coverage.out

.github/workflows/pr-codespell.yml

@@ -8,14 +8,14 @@ jobs:
     steps:
 
     - name: Check out the code
-      uses: actions/checkout@v4
+      uses: actions/checkout@v6
 
     - name: Codespell
       uses: codespell-project/actions-codespell@master
       with:
-        # ignore the config/.../crd.go file as it's generated binary data that is edited elswhere.
+        # ignore the config/.../crd.go file as it's generated binary data that is edited elsewhere.
         skip: .git,*.png,*.jpg,*.woff,*.ttf,*.gif,*.ico,./config/crd/v1beta1/crds/crds.go,./config/crd/v1/crds/crds.go,./config/crd/v2alpha1/crds/crds.go,./go.sum,./LICENSE
-        ignore_words_list: iam,aks,ist,bridget,ue,shouldnot,atleast
+        ignore_words_list: iam,aks,ist,bridget,ue,shouldnot,atleast,notin,sme,optin,sie
         check_filenames: true
         check_hidden: true
 

.github/workflows/pr-containers.yml

@@ -13,7 +13,7 @@ jobs:
     name: Build
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v6
       name: Checkout
 
     - name: Set up QEMU

.github/workflows/pr-filepath-check.yml

@@ -0,0 +1,93 @@
+name: Pull Request File Path Check
+on: [pull_request]
+jobs:
+
+  filepath-check:
+    name: Check for invalid characters in file paths
+    runs-on: ubuntu-latest
+    steps:
+
+    - name: Check out the code
+      uses: actions/checkout@v6
+
+    - name: Validate file paths for Go module compatibility
+      run: |
+        # Go's module zip rejects filenames containing certain characters.
+        # See golang.org/x/mod/module fileNameOK() for the full specification.
+        #
+        # Allowed ASCII: letters, digits, and: !#$%&()+,-.=@[]^_{}~ and space
+        # Allowed non-ASCII: unicode letters only
+        # Rejected: " ' * < > ? ` | / \ : and any non-letter unicode (control
+        #           chars, format chars like U+200E LEFT-TO-RIGHT MARK, etc.)
+        #
+        # This check catches issues like the U+200E incident in PR #9552.
+
+        EXIT_STATUS=0
+
+        git ls-files -z | python3 -c "
+        import sys, unicodedata
+
+        data = sys.stdin.buffer.read()
+        files = data.split(b'\x00')
+
+        # Characters explicitly rejected by Go's fileNameOK
+        # (path separators / and \ are inherent to paths so we check per-element)
+        bad_ascii = set('\"' + \"'\" + '*<>?\`|:')
+
+        allowed_ascii = set('!#$%&()+,-.=@[]^_{}~ ')
+
+        def is_ok(ch):
+            if ch.isascii():
+                return ch.isalnum() or ch in allowed_ascii
+            return ch.isalpha()
+
+        bad_files = []  # list of (original_path, clean_path, char_desc)
+        for f in files:
+            if not f:
+                continue
+            try:
+                name = f.decode('utf-8')
+            except UnicodeDecodeError:
+                print(f'::error::Non-UTF-8 bytes in filename: {f!r}')
+                bad_files.append((repr(f), None, 'non-UTF-8 bytes'))
+                continue
+
+            # Check each path element (split on /)
+            for element in name.split('/'):
+                for ch in element:
+                    if not is_ok(ch):
+                        cp = ord(ch)
+                        char_name = unicodedata.name(ch, f'U+{cp:04X}')
+                        char_desc = f'U+{cp:04X} ({char_name})'
+                        # Build cleaned path by stripping invalid chars
+                        clean = '/'.join(
+                            ''.join(c for c in elem if is_ok(c))
+                            for elem in name.split('/')
+                        )
+                        print(f'::error file={name}::File \"{name}\" contains invalid char {char_desc}')
+                        bad_files.append((name, clean, char_desc))
+                        break
+
+        if bad_files:
+            print()
+            print('The following files have characters that are invalid in Go module zip archives:')
+            print()
+            for original, clean, desc in bad_files:
+                print(f'  {original}  — {desc}')
+            print()
+            print('To fix, rename the files to remove the problematic characters:')
+            print()
+            for original, clean, desc in bad_files:
+                if clean:
+                    print(f'  mv \"{original}\" \"{clean}\" && git add \"{clean}\"')
+                    print(f'  # or: git mv \"{original}\" \"{clean}\"')
+                else:
+                    print(f'  # {original} — cannot auto-suggest rename (non-UTF-8)')
+            print()
+            print('See https://github.com/vmware-tanzu/velero/pull/9552 for context.')
+            sys.exit(1)
+        else:
+            print('All file paths are valid for Go module zip.')
+        " || EXIT_STATUS=1
+
+        exit $EXIT_STATUS

.github/workflows/pr-goreleaser.yml

@@ -14,7 +14,7 @@ jobs:
     name: Build
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v6
       name: Checkout
 
     - name: Verify .goreleaser.yml and try a dryrun release.

.github/workflows/pr-linter-check.yml

@@ -1,20 +1,32 @@
 name: Pull Request Linter Check
-on: [pull_request]
+on:
+  pull_request:
+    # Do not run when the change only includes these directories.
+    paths-ignore:
+      - "site/**"
+      - "design/**"
+      - "**/*.md"
 jobs:
+  get-go-version:
+    uses: ./.github/workflows/get-go-version.yaml
+    with:
+      ref: ${{ github.event.pull_request.base.ref }}
 
   build:
     name: Run Linter Check
     runs-on: ubuntu-latest
+    needs: get-go-version
     steps:
       - name: Check out the code
-        uses: actions/checkout@v4
-      - name: Set up Go
-        uses: actions/setup-go@v5
+        uses: actions/checkout@v6
+
+      - name: Set up Go version
+        uses: actions/setup-go@v6
         with:
-          go-version-file: 'go.mod'
-        id: go
+          go-version: ${{ needs.get-go-version.outputs.version }}
+
       - name: Linter check
-        uses: golangci/golangci-lint-action@v6
+        uses: golangci/golangci-lint-action@v9
         with:
-          version: v1.57.2
+          version: v2.5.0
           args: --verbose

.github/workflows/push-builder.yml

@@ -12,7 +12,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
 
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v6
       with:
         # The default value is "1" which fetches only a single commit. If we merge PR without squash or rebase,
         # there are at least two commits: the first one is the merge commit and the second one is the real commit

.github/workflows/push.yml

@@ -9,100 +9,55 @@ on:
       - '*'
 
 jobs:
+  get-go-version:
+    uses: ./.github/workflows/get-go-version.yaml
+    with:
+      ref: ${{ github.ref_name }}
 
   build:
     name: Build
     runs-on: ubuntu-latest
+    needs: get-go-version
     steps:
+      - name: Check out the code
+        uses: actions/checkout@v6
 
-    - name: Set up Go
-      uses: actions/setup-go@v5
-      with:
-        go-version: '1.22'
-      id: go
+      - name: Set up Go version
+        uses: actions/setup-go@v6
+        with:
+          go-version: ${{ needs.get-go-version.outputs.version }}
 
-    - uses: actions/checkout@v4
-
-    # Fix issue of setup-gcloud
-    - run: |
-        sudo apt-get install python2.7
-        export CLOUDSDK_PYTHON="/usr/bin/python2"
-
-    - id: 'auth'
-      uses: google-github-actions/auth@v2
-      with:
-        credentials_json: '${{ secrets.GCS_SA_KEY }}'
-
-    - name: 'set up GCloud SDK'
-      uses: google-github-actions/setup-gcloud@v2
-
-    - name: 'use gcloud CLI'
-      run: |
-        gcloud info
-
-    - name: Set up QEMU
-      id: qemu
-      uses: docker/setup-qemu-action@v3
-      with:
-        platforms: all
-
-    - name: Set up Docker Buildx
-      id: buildx
-      uses: docker/setup-buildx-action@v3
-      with:
-        version: latest
-
-    - name: Build
-      run: |
-        make local
-        # Clean go cache to ease the build environment storage pressure.
-        go clean -modcache -cache
-
-    - name: Test
-      run: make test
-
-    - name: Upload test coverage
-      uses: codecov/codecov-action@v4
-      with:
-        token: ${{ secrets.CODECOV_TOKEN }}
-        files: coverage.out
-        verbose: true
-
-    # Use the JSON key in secret to login gcr.io
-    - uses: 'docker/login-action@v3'
-      with:
-        registry: 'gcr.io' # or REGION.docker.pkg.dev
-        username: '_json_key'
-        password: '${{ secrets.GCR_SA_KEY }}'
-
-    # Only try to publish the container image from the root repo; forks don't have permission to do so and will always get failures.
-    - name: Publish container image
-      if: github.repository == 'vmware-tanzu/velero'
-      run: |
-        sudo swapoff -a
-        sudo rm -f /mnt/swapfile
-        docker system prune -a --force
+      - name: Set up QEMU
+        id: qemu
+        uses: docker/setup-qemu-action@v3
+        with:
+          platforms: all
+      - name: Set up Docker Buildx
+        id: buildx
+        uses: docker/setup-buildx-action@v3
+        with:
+          version: latest
+      - name: Build
+        run: |
+          make local
+          # Clean go cache to ease the build environment storage pressure.
+          go clean -modcache -cache
+      - name: Test
+        run: make test
+      - name: Upload test coverage
+        uses: codecov/codecov-action@v5
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          files: coverage.out
+          verbose: true
+      # Only try to publish the container image from the root repo; forks don't have permission to do so and will always get failures.
+      - name: Publish container image
+        if: github.repository == 'vmware-tanzu/velero'
+        run: |
+          sudo swapoff -a
+          sudo rm -f /mnt/swapfile
+          docker system prune -a --force
               
-        # Build and push Velero image to docker registry
-        docker login -u ${{ secrets.DOCKER_USER }} -p ${{ secrets.DOCKER_PASSWORD }}
-        VERSION=$(./hack/docker-push.sh | grep 'VERSION:' | awk -F: '{print $2}' | xargs)
-
-        # Upload Velero image package to GCS
-        source hack/ci/build_util.sh
-        BIN=velero
-        RESTORE_HELPER_BIN=velero-restore-helper
-        GCS_BUCKET=velero-builds
-        VELERO_IMAGE=${BIN}-${VERSION}
-        VELERO_RESTORE_HELPER_IMAGE=${RESTORE_HELPER_BIN}-${VERSION}
-        VELERO_IMAGE_FILE=${VELERO_IMAGE}.tar.gz
-        VELERO_RESTORE_HELPER_IMAGE_FILE=${VELERO_RESTORE_HELPER_IMAGE}.tar.gz
-        VELERO_IMAGE_BACKUP_FILE=${VELERO_IMAGE}-'build.'${GITHUB_RUN_NUMBER}.tar.gz
-        VELERO_RESTORE_HELPER_IMAGE_BACKUP_FILE=${VELERO_RESTORE_HELPER_IMAGE}-'build.'${GITHUB_RUN_NUMBER}.tar.gz
-
-        cp ${VELERO_IMAGE_FILE} ${VELERO_IMAGE_BACKUP_FILE}
-        cp ${VELERO_RESTORE_HELPER_IMAGE_FILE} ${VELERO_RESTORE_HELPER_IMAGE_BACKUP_FILE}
-
-        uploader ${VELERO_IMAGE_FILE} ${GCS_BUCKET}
-        uploader ${VELERO_RESTORE_HELPER_IMAGE_FILE} ${GCS_BUCKET}
-        uploader ${VELERO_IMAGE_BACKUP_FILE} ${GCS_BUCKET}
-        uploader ${VELERO_RESTORE_HELPER_IMAGE_BACKUP_FILE} ${GCS_BUCKET}
+          # Build and push Velero image to docker registry
+          docker login -u ${{ secrets.DOCKER_USER }} -p ${{ secrets.DOCKER_PASSWORD }}
+          ./hack/docker-push.sh

.github/workflows/rebase.yml

@@ -9,7 +9,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Checkout the latest code
-      uses: actions/checkout@v4
+      uses: actions/checkout@v6
       with:
         fetch-depth: 0
     - name: Automatic Rebase

.github/workflows/stale-issues.yml

@@ -7,7 +7,7 @@ jobs:
   stale:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/stale@v9.0.0
+      - uses: actions/stale@v10.1.1
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
           stale-issue-message: "This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 14 days. If a Velero team member has requested log or more information, please provide the output of the shared commands."
@@ -20,4 +20,4 @@ jobs:
           days-before-pr-close: -1
           # Only issues made after Feb 09 2021.
           start-date: "2021-09-02T00:00:00"
-          exempt-issue-labels: "Epic,Area/CLI,Area/Cloud/AWS,Area/Cloud/Azure,Area/Cloud/GCP,Area/Cloud/vSphere,Area/CSI,Area/Design,Area/Documentation,Area/Plugins,Bug,Enhancement/User,kind/requirement,kind/refactor,kind/tech-debt,limitation,Needs investigation,Needs triage,Needs Product,P0 - Hair on fire,P1 - Important,P2 - Long-term important,P3 - Wouldn't it be nice if...,Product Requirements,Restic - GA,Restic,release-blocker,Security"
+          exempt-issue-labels: "Epic,Area/CLI,Area/Cloud/AWS,Area/Cloud/Azure,Area/Cloud/GCP,Area/Cloud/vSphere,Area/CSI,Area/Design,Area/Documentation,Area/Plugins,Bug,Enhancement/User,kind/requirement,kind/refactor,kind/tech-debt,limitation,Needs investigation,Needs triage,Needs Product,P0 - Hair on fire,P1 - Important,P2 - Long-term important,P3 - Wouldn't it be nice if...,Product Requirements,Restic - GA,Restic,release-blocker,Security,backlog"

.gitignore

@@ -53,4 +53,13 @@ tilt-resources/cloud
 # test generated files
 test/e2e/report.xml
 coverage.out
-__debug_bin*
+__debug_bin*
+debug.test*
+
+# make lint cache
+.cache/
+
+# Go telemetry directory created when container sets HOME to working directory
+# This happens because Makefile uses 'docker run -w /github.com/vmware-tanzu/velero'
+# and Go's os.UserConfigDir() falls back to $HOME/.config when XDG_CONFIG_HOME is unset
+.config/

.golangci.yaml

@@ -6,17 +6,12 @@ run:
   # default concurrency is a available CPU number
   concurrency: 4
 
-  # timeout for analysis, e.g. 30s, 5m, default is 1m
+  # timeout for analysis, e.g. 30s, 5m, default is 0
   timeout: 20m
 
   # exit code when at least one issue was found, default is 1
   issues-exit-code: 1
 
-
-  # default is true. Enables skipping of directories:
-  #   vendor$, third_party$, testdata$, examples$, Godeps$, builtin$
-  skip-dirs-use-default: true
-
   # by default isn't set. If set we pass it to "go list -mod={option}". From "go help modules":
   # If invoked with -mod=readonly, the go command is disallowed from the implicit
   # automatic updating of go.mod described above. Instead, it fails when any changes
@@ -32,378 +27,403 @@ run:
   # If false (default) - golangci-lint acquires file lock on start.
   allow-parallel-runners: false
 
-
 # output configuration options
 output:
-  # colored-line-number|line-number|json|tab|checkstyle|code-climate, default is "colored-line-number"
   formats:
-    - format: colored-line-number
+    text:
       path: stdout
 
-  # print lines of code with issue, default is true
-  print-issued-lines: true
+      # print lines of code with issue, default is true
+      print-issued-lines: true
 
-  # print linter name in the end of issue text, default is true
-  print-linter-name: true
+      # print linter name in the end of issue text, default is true
+      print-linter-name: true
 
-  # make issues output unique by line, default is true
-  uniq-by-line: true
-
-
-# all available settings of specific linters
-linters-settings:
-  dogsled:
-    # checks assignments with too many blank identifiers; default is 2
-    max-blank-identifiers: 2
-  dupl:
-    # tokens count to trigger issue, 150 by default
-    threshold: 100
-  errcheck:
-    # report about not checking of errors in type assertions: `a := b.(MyStruct)`;
-    # default is false: such cases aren't reported by default.
-    check-type-assertions: false
-
-    # report about assignment of errors to blank identifier: `num, _ := strconv.Atoi(numStr)`;
-    # default is false: such cases aren't reported by default.
-    check-blank: false
-
-    # [deprecated] comma-separated list of pairs of the form pkg:regex
-    # the regex is used to ignore names within pkg. (default "fmt:.*").
-    # see https://github.com/kisielk/errcheck#the-deprecated-method for details
-    # ignore: fmt:.*,io/ioutil:^Read.*
-
-    # path to a file containing a list of functions to exclude from checking
-    # see https://github.com/kisielk/errcheck#excluding-functions for details
-    # exclude: /path/to/file.txt
-  exhaustive:
-    # indicates that switch statements are to be considered exhaustive if a
-    # 'default' case is present, even if all enum members aren't listed in the
-    # switch
-    default-signifies-exhaustive: false
-  funlen:
-    lines: 60
-    statements: 40
-  gocognit:
-    # minimal code complexity to report, 30 by default (but we recommend 10-20)
-    min-complexity: 10
-  nestif:
-    # minimal complexity of if statements to report, 5 by default
-    min-complexity: 4
-  goconst:
-    # minimal length of string constant, 3 by default
-    min-len: 3
-    # minimal occurrences count to trigger, 3 by default
-    min-occurrences: 5
-  gocritic:
-    # Which checks should be enabled; can't be combined with 'disabled-checks';
-    # See https://go-critic.github.io/overview#checks-overview
-    # To check which checks are enabled run `GL_DEBUG=gocritic golangci-lint run`
-    # By default list of stable checks is used.
-    # enabled-checks:
-    #  - rangeValCopy
-
-    # Which checks should be disabled; can't be combined with 'enabled-checks'; default is empty
-    # disabled-checks:
-    #  - regexpMust
-
-    # Enable multiple checks by tags, run `GL_DEBUG=gocritic golangci-lint run` to see all tags and checks.
-    # Empty list by default. See https://github.com/go-critic/go-critic#usage -> section "Tags".
-    # enabled-tags:
-    #  - performance
-    # disabled-tags:
-    #  - experimental
-
-    settings: # settings passed to gocritic
-      captLocal: # must be valid enabled check name
-        paramsOnly: true
-    #  rangeValCopy:
-    #    sizeThreshold: 32
-  gocyclo:
-    # minimal code complexity to report, 30 by default (but we recommend 10-20)
-    min-complexity: 10
-  godot:
-    # check all top-level comments, not only declarations
-    check-all: false
-  godox:
-    # report any comments starting with keywords, this is useful for TODO or FIXME comments that
-    # might be left in the code accidentally and should be resolved before merging
-    keywords: # default keywords are TODO, BUG, and FIXME, these can be overwritten by this setting
-      - NOTE
-      - OPTIMIZE # marks code that should be optimized before merging
-      - HACK # marks hack-arounds that should be removed before merging
-  gofmt:
-    # simplify code: gofmt with `-s` option, true by default
-    simplify: true
-  goimports:
-    # put imports beginning with prefix after 3rd-party packages;
-    # it's a comma-separated list of prefixes
-    local-prefixes: github.com/org/project
-  golint:
-    # minimal confidence for issues, default is 0.8
-    min-confidence: 0.8
-  gomnd:
-    # the list of enabled checks, see https://github.com/tommy-muehle/go-mnd/#checks for description.
-    checks: argument,case,condition,operation,return,assign
-  gomodguard:
-    allowed:
-      modules:                                                        # List of allowed modules
-        # - gopkg.in/yaml.v2
-      domains:                                                        # List of allowed module domains
-        # - golang.org
-    blocked:
-      modules:                                                        # List of blocked modules
-        # - github.com/uudashr/go-module:                             # Blocked module
-        #     recommendations:                                        # Recommended modules that should be used instead (Optional)
-        #       - golang.org/x/mod
-        #     reason: "`mod` is the official go.mod parser library."  # Reason why the recommended module should be used (Optional)
-      versions:                                                       # List of blocked module version constraints
-        # - github.com/mitchellh/go-homedir:                          # Blocked module with version constraint
-        #     version: "< 1.1.0"                                      # Version constraint, see https://github.com/Masterminds/semver#basic-comparisons
-        #     reason: "testing if blocked version constraint works."  # Reason why the version constraint exists. (Optional)
-  govet:
-    # report about shadowed variables
-    # check-shadowing: true
-
-    # settings per analyzer
-    settings:
-      printf: # analyzer name, run `go tool vet help` to see all analyzers
-        funcs: # run `go tool vet help printf` to see available settings for `printf` analyzer
-          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Infof
-          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Warnf
-          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Errorf
-          - (github.com/golangci/golangci-lint/pkg/logutils.Log).Fatalf
-
-    # enable or disable analyzers by name
-    enable:
-      - atomicalign
-    enable-all: false
-    disable:
-      - shadow
-    disable-all: false
-  depguard:
-    list-type: blacklist # Velero.io word list : ignore
-    include-go-root: false
-    packages:
-      - github.com/sirupsen/logrus
-    packages-with-error-message:
-      # specify an error message to output when a denylisted package is used
-      - github.com/sirupsen/logrus: "logging is allowed only by logutils.Log"
-  lll:
-    # max line length, lines longer will be reported. Default is 120.
-    # '\t' is counted as 1 character by default, and can be changed with the tab-width option
-    line-length: 120
-    # tab width in spaces. Default to 1.
-    tab-width: 1
-  maligned:
-    # print struct with more effective memory layout or not, false by default
-    suggest-new: true
-  misspell:
-    # Correct spellings using locale preferences for US or UK.
-    # Default is to use a neutral variety of English.
-    # Setting locale to US will correct the British spelling of 'colour' to 'color'.
-    locale: US
-    ignore-words:
-      - someword
-  nakedret:
-    # make an issue if func has more lines of code than this setting and it has naked returns; default is 30
-    max-func-lines: 30
-  prealloc:
-    # XXX: we don't recommend using this linter before doing performance profiling.
-    # For most programs usage of prealloc will be a premature optimization.
-
-    # Report preallocation suggestions only on simple loops that have no returns/breaks/continues/gotos in them.
-    # True by default.
-    simple: true
-    range-loops: true # Report preallocation suggestions on range loops, true by default
-    for-loops: false # Report preallocation suggestions on for loops, false by default
-  nolintlint:
-    # Enable to ensure that nolint directives are all used. Default is true.
-    allow-unused: false
-    # Disable to ensure that nolint directives don't have a leading space. Default is true.
-    allow-leading-space: true
-    # Exclude following linters from requiring an explanation.  Default is [].
-    allow-no-explanation: []
-    # Enable to require an explanation of nonzero length after each nolint directive. Default is false.
-    require-explanation: true
-    # Enable to require nolint directives to mention the specific linter being suppressed. Default is false.
-    require-specific: true
-  revive:
-    rules:
-      - name: unexported-return
-        disabled: true
-    
-  rowserrcheck:
-    packages:
-      - github.com/jmoiron/sqlx
-  testifylint:
-      # TODO: enable them all
-      disable:
-        - bool-compare
-        - compares
-        - error-is-as
-        - error-nil
-        - expected-actual
-        - go-require
-        - float-compare
-        - require-error
-        - suite-dont-use-pkg
-        - suite-extra-assert-call
-        - suite-thelper
-      enable:
-        - empty
-        - len
-        - nil-compare
-  testpackage:
-    # regexp pattern to skip files
-    skip-regexp: (export|internal)_test\.go
-  unparam:
-    # Inspect exported functions, default is false. Set to true if no external program/library imports your code.
-    # XXX: if you enable this setting, unparam will report a lot of false-positives in text editors:
-    # if it's called for subdir of a project it can't find external interfaces. All text editor integrations
-    # with golangci-lint call it on a directory with the changed file.
-    check-exported: false
-  unused:
-    # treat code as a program (not a library) and report unused exported identifiers; default is false.
-    # XXX: if you enable this setting, unused will report a lot of false-positives in text editors:
-    # if it's called for subdir of a project it can't find funcs usages. All text editor integrations
-    # with golangci-lint call it on a directory with the changed file.
-    check-exported: false
-  whitespace:
-    multi-if: false   # Enforces newlines (or comments) after every multi-line if statement
-    multi-func: false # Enforces newlines (or comments) after every multi-line function signature
-  wsl:
-    # If true append is only allowed to be cuddled if appending value is
-    # matching variables, fields or types on line above. Default is true.
-    strict-append: true
-    # Allow calls and assignments to be cuddled as long as the lines have any
-    # matching variables, fields or types. Default is true.
-    allow-assign-and-call: true
-    # Allow multiline assignments to be cuddled. Default is true.
-    allow-multiline-assign: true
-    # Allow declarations (var) to be cuddled.
-    allow-cuddle-declarations: false
-    # Allow trailing comments in ending of blocks
-    allow-trailing-comment: false
-    # Force newlines in end of case at this limit (0 = never).
-    force-case-trailing-whitespace: 0
-    # Force cuddling of err checks with err var assignment
-    force-err-cuddling: false
-    # Allow leading comments to be separated with empty liens
-    allow-separated-leading-comment: false
+  # Show statistics per linter.      
+  show-stats: false
 
 linters:
-  disable-all: true
+  # all available settings of specific linters
+  settings:
+    depguard:
+      rules:
+        main:
+          deny:
+            # specify an error message to output when a denylisted package is used
+            - pkg: github.com/sirupsen/logrus
+              desc: "logging is allowed only by logutils.Log"
+
+    dogsled:
+      # checks assignments with too many blank identifiers; default is 2
+      max-blank-identifiers: 2
+
+    dupl:
+      # tokens count to trigger issue, 150 by default
+      threshold: 100
+
+    errcheck:
+      # report about not checking of errors in type assertions: `a := b.(MyStruct)`;
+      # default is false: such cases aren't reported by default.
+      check-type-assertions: false
+
+      # report about assignment of errors to blank identifier: `num, _ := strconv.Atoi(numStr)`;
+      # default is false: such cases aren't reported by default.
+      check-blank: false
+
+
+    exhaustive:
+      # indicates that switch statements are to be considered exhaustive if a
+      # 'default' case is present, even if all enum members aren't listed in the
+      # switch
+      default-signifies-exhaustive: false
+
+    funlen:
+      lines: 60
+      statements: 40
+
+    gocognit:
+      # minimal code complexity to report, 30 by default (but we recommend 10-20)
+      min-complexity: 10
+
+    nestif:
+      # minimal complexity of if statements to report, 5 by default
+      min-complexity: 4
+
+    goconst:
+      # minimal length of string constant, 3 by default
+      min-len: 3
+      # minimal occurrences count to trigger, 3 by default
+      min-occurrences: 5
+
+    gocritic:
+      # Which checks should be enabled; can't be combined with 'disabled-checks';
+      # See https://go-critic.github.io/overview#checks-overview
+      # To check which checks are enabled run `GL_DEBUG=gocritic golangci-lint run`
+      # By default list of stable checks is used.
+      settings: # settings passed to gocritic
+        captLocal: # must be valid enabled check name
+          paramsOnly: true
+
+    gocyclo:
+      # minimal code complexity to report, 30 by default (but we recommend 10-20)
+      min-complexity: 10
+
+    godot:
+      # check all top-level comments, not only declarations
+      check-all: false
+
+    godox:
+      # report any comments starting with keywords, this is useful for TODO or FIXME comments that
+      # might be left in the code accidentally and should be resolved before merging
+      keywords: # default keywords are TODO, BUG, and FIXME, these can be overwritten by this setting
+        - NOTE
+        - OPTIMIZE # marks code that should be optimized before merging
+        - HACK # marks hack-arounds that should be removed before merging
+
+    gosec:
+      excludes:
+        - G115
+
+    govet:
+      # enable or disable analyzers by name
+      enable:
+        - atomicalign
+      enable-all: false
+      disable:
+        - shadow
+      disable-all: false
+  
+    importas:
+       alias:
+        - alias: appsv1api
+          pkg: k8s.io/api/apps/v1
+        - alias: corev1api
+          pkg: k8s.io/api/core/v1
+        - alias: rbacv1
+          pkg: k8s.io/api/rbac/v1
+        - alias: apierrors
+          pkg: k8s.io/apimachinery/pkg/api/errors
+        - alias: apiextv1
+          pkg: k8s.io/apiextensions-apiserver/pkg/apis/apiextensions/v1
+        - alias: metav1
+          pkg: k8s.io/apimachinery/pkg/apis/meta/v1
+        - alias: storagev1api
+          pkg: k8s.io/api/storage/v1
+        - alias: batchv1api
+          pkg: k8s.io/api/batch/v1
+
+    lll:
+    # max line length, lines longer will be reported. Default is 120.
+    # '\t' is counted as 1 character by default, and can be changed with the tab-width option
+      line-length: 120
+      # tab width in spaces. Default to 1.
+      tab-width: 1
+
+    misspell:
+      # Correct spellings using locale preferences for US or UK.
+      # Default is to use a neutral variety of English.
+      # Setting locale to US will correct the British spelling of 'colour' to 'color'.
+      locale: US
+      ignore-rules:
+        - someword
+
+    nakedret:
+      # make an issue if func has more lines of code than this setting and it has naked returns; default is 30
+      max-func-lines: 30
+
+    prealloc:
+      # XXX: we don't recommend using this linter before doing performance profiling.
+      # For most programs usage of prealloc will be a premature optimization.
+
+      # Report preallocation suggestions only on simple loops that have no returns/breaks/continues/gotos in them.
+      # True by default.
+      simple: true
+      range-loops: true # Report preallocation suggestions on range loops, true by default
+      for-loops: false # Report preallocation suggestions on for loops, false by default
+
+    nolintlint:
+      # Enable to ensure that nolint directives are all used. Default is true.
+      allow-unused: false
+      # Exclude following linters from requiring an explanation.  Default is [].
+      allow-no-explanation: []
+      # Enable to require an explanation of nonzero length after each nolint directive. Default is false.
+      require-explanation: true
+      # Enable to require nolint directives to mention the specific linter being suppressed. Default is false.
+      require-specific: true
+
+    perfsprint:
+      strconcat: false
+      sprintf1: false
+      errorf: false
+      int-conversion: true
+
+    revive:
+      rules:
+        - name: blank-imports
+          disabled: true
+        - name: context-as-argument
+          disabled: true
+        - name: context-keys-type
+        - name: dot-imports
+          disabled: true
+        - name: early-return
+          disabled: true
+          arguments:
+            - "preserveScope"
+        - name: empty-block
+          disabled: true
+        - name: error-naming
+          disabled: true
+        - name: error-return
+          disabled: true
+        - name: error-strings
+          disabled: true
+        - name: errorf
+          disabled: true
+        - name: increment-decrement
+        - name: indent-error-flow
+          disabled: true
+        - name: range
+        - name: receiver-naming
+          disabled: true
+        - name: redefines-builtin-id
+          disabled: true
+        - name: superfluous-else
+          disabled: true
+          arguments:
+            - "preserveScope"
+        - name: time-naming
+        - name: unexported-return
+          disabled: true
+        - name: unnecessary-stmt
+        - name: unreachable-code
+        - name: unused-parameter
+          disabled: true
+        - name: use-any
+        - name: var-declaration
+        - name: var-naming
+          disabled: true
+
+    rowserrcheck:
+      packages:
+        - github.com/jmoiron/sqlx
+
+    staticcheck:
+      checks:
+        - all
+        - -QF1001 # FIXME
+        - -QF1003 # FIXME
+        - -QF1004 # FIXME
+        - -QF1007 # FIXME
+        - -QF1008 # FIXME
+        - -QF1009 # FIXME
+        - -QF1012 # FIXME
+
+    testifylint:
+      # TODO: enable them all
+      disable:
+        - float-compare
+        - go-require
+      enable-all: true
+
+    testpackage:
+      # regexp pattern to skip files
+      skip-regexp: (export|internal)_test\.go
+    unparam:
+      # Inspect exported functions, default is false. Set to true if no external program/library imports your code.
+      # XXX: if you enable this setting, unparam will report a lot of false-positives in text editors:
+      # if it's called for subdir of a project it can't find external interfaces. All text editor integrations
+      # with golangci-lint call it on a directory with the changed file.
+      check-exported: false
+
+    usetesting:
+      os-setenv: false
+
+    whitespace:
+      multi-if: false # Enforces newlines (or comments) after every multi-line if statement
+      multi-func: false # Enforces newlines (or comments) after every multi-line function signature
+
+    wsl:
+      # If true append is only allowed to be cuddled if appending value is
+      # matching variables, fields or types on line above. Default is true.
+      strict-append: true
+      # Allow calls and assignments to be cuddled as long as the lines have any
+      # matching variables, fields or types. Default is true.
+      allow-assign-and-call: true
+      # Allow multiline assignments to be cuddled. Default is true.
+      allow-multiline-assign: true
+      # Allow declarations (var) to be cuddled.
+      allow-cuddle-declarations: false
+      # Allow trailing comments in ending of blocks
+      allow-trailing-comment: false
+      # Force newlines in end of case at this limit (0 = never).
+      force-case-trailing-whitespace: 0
+      # Force cuddling of err checks with err var assignment
+      force-err-cuddling: false
+      # Allow leading comments to be separated with empty lines
+      allow-separated-leading-comment: false
+
+  default: none
   enable:
     - asasalint
     - asciicheck
     - bidichk
     - bodyclose
+    - copyloopvar
     - dogsled
-    - durationcheck
     - dupword
+    - durationcheck
     - errcheck
-    - exportloopref
     - errchkjson
+    - exptostd
+    - ginkgolinter
     - goconst
-    - gofmt
     - goheader
-    - goimports
     - goprintffuncname
     - gosec
-    - gosimple
     - govet
-    - ginkgolinter
     - importas
     - ineffassign
     - misspell
     - nakedret
-    - nosprintfhostport
     - nilerr
     - noctx
     - nolintlint
+    - nosprintfhostport
+    - perfsprint
     - revive
     - staticcheck
-    - stylecheck
     - testifylint
-    - typecheck
+    - thelper
     - unconvert
     - unparam
     - unused
     - usestdlibvars
+    - usetesting
     - whitespace
-  fast: false
 
+  exclusions:
+    # which dirs to skip: issues from them won't be reported;
+    # can use regexp here: generated.*, regexp is applied on full path;
+    # default value is empty list, but default dirs are skipped independently
+    # from this option's value (see skip-dirs-use-default).
+    # "/" will be replaced by current OS file path separator to properly work
+    # on Windows.
+    paths:
+      - pkg/plugin/generated/*
+      - third_party
+
+    rules:
+      - linters:
+          - staticcheck
+        text: "DefaultVolumesToRestic" # No need to report deprecate for DefaultVolumesToRestic.
+      - path: ".*_test.go$"
+        linters:
+          - errcheck
+          - goconst
+          - gosec
+          - govet
+          - staticcheck
+          - unparam
+          - unused
+      - path: test/
+        linters:
+          - errcheck
+          - goconst
+          - gosec
+          - nilerr
+          - staticcheck
+          - unparam
+          - unused
+      - path: ".*data_upload_controller_test.go$"
+        linters:
+          - dupword
+        text: "type"
+      - path: ".*config_test.go$"
+        linters:
+          - dupword
+        text: "bucket"
+
+    generated: lax
+    presets:
+      - comments
+      - common-false-positives
+      - legacy
+      - std-error-handling
 
 issues:
-  exclude-rules:
-    - linters:
-        - staticcheck
-      text: "github.com/golang/protobuf/proto" # grpc-go still uses github.com/golang/protobuf/proto.
-    - linters:
-        - staticcheck
-      text: "DefaultVolumesToRestic" # No need to report deprecate for DefaultVolumesToRestic.
-    - path: ".*_test.go$"
-      linters:
-        - dupword
-        - errcheck
-        - goconst
-        - gosec
-        - govet
-        - staticcheck
-        - stylecheck
-        - unconvert
-        - unparam
-        - unused
-    - path: test/
-      linters:
-        - dupword
-        - errcheck
-        - goconst
-        - gosec
-        - gosimple
-        - nilerr
-        - staticcheck
-        - stylecheck
-        - unconvert
-        - unparam
-        - unused
-
-  # The list of ids of default excludes to include or disable. By default it's empty.
-  include:
-    - EXC0002 # disable excluding of issues about comments from golint
-
   # Maximum issues count per one linter. Set to 0 to disable. Default is 50.
   max-issues-per-linter: 0
 
   # Maximum count of issues with the same text. Set to 0 to disable. Default is 3.
   max-same-issues: 0
 
-  # Show only new issues created after git revision `REV`
-  # new-from-rev: origin/main
+  # make issues output unique by line, default is true
+  uniq-by-line: true
 
-  # which dirs to skip: issues from them won't be reported;
-  # can use regexp here: generated.*, regexp is applied on full path;
-  # default value is empty list, but default dirs are skipped independently
-  # from this option's value (see skip-dirs-use-default).
-  # "/" will be replaced by current OS file path separator to properly work
-  # on Windows.
-  exclude-dirs:
-    - pkg/plugin/generated/*
+# This file contains all available configuration options
+# with their default values.
+formatters:
+  enable:
+    - gofmt
+    - goimports
+
+  exclusions:
+    generated: lax
+    paths:
+      - pkg/plugin/generated/*
+      - third_party
+
+  settings:
+    gofmt:
+      # simplify code: gofmt with `-s` option, true by default
+      simplify: true
+    goimports:
+      local-prefixes:
+        - github.com/vmware-tanzu/velero
 
 severity:
-  # Default value is empty string.
-  # Set the default severity for issues. If severity rules are defined and the issues 
-  # do not match or no severity is provided to the rule this will be the default 
-  # severity applied. Severities should match the supported severity names of the 
-  # selected out format.
-  # - Code climate: https://docs.codeclimate.com/docs/issues#issue-severity
-  # -   Checkstyle: https://checkstyle.sourceforge.io/property_types.html#severity
-  # -       Github: https://help.github.com/en/actions/reference/workflow-commands-for-github-actions#setting-an-error-message
-  default-severity: error
-
-  # The default value is false. 
-  # If set to true severity-rules regular expressions become case sensitive.
-  case-sensitive: false
+  default: error
 
   # Default value is empty list.
   # When a list of severity rules are provided, severity information will be added to lint
@@ -412,5 +432,7 @@ severity:
   # Only affects out formats that support setting severity information.
   rules:
     - linters:
-      - dupl
+        - dupl
       severity: info
+
+version: "2"

.goreleaser.yml

@@ -26,18 +26,23 @@ builds:
       - arm
       - arm64
       - ppc64le
+      - s390x
     ignore:
       # don't build arm for darwin and arm/arm64 for windows
       - goos: darwin
         goarch: arm
       - goos: darwin
         goarch: ppc64le
+      - goos: darwin
+        goarch: s390x
       - goos: windows
         goarch: arm
       - goos: windows
         goarch: arm64
       - goos: windows
         goarch: ppc64le
+      - goos: windows
+        goarch: s390x
     ldflags:
       - -X "github.com/vmware-tanzu/velero/pkg/buildinfo.Version={{ .Tag }}" -X "github.com/vmware-tanzu/velero/pkg/buildinfo.GitSHA={{ .FullCommit }}" -X "github.com/vmware-tanzu/velero/pkg/buildinfo.GitTreeState={{ .Env.GIT_TREE_STATE }}" -X "github.com/vmware-tanzu/velero/pkg/buildinfo.ImageRegistry={{ .Env.REGISTRY }}"
 archives:
@@ -46,9 +51,6 @@ archives:
     files:
       - LICENSE
       - examples/**/*
-    # Add the setting to resolve the DEPRECATED warning. Actually, Velero's case is not affected by the rlcp behavior change.
-    # https://github.com/orgs/goreleaser/discussions/3659#discussioncomment-4587257
-    rlcp: true 
 checksum:
   name_template: 'CHECKSUM'
 release:
@@ -63,4 +65,4 @@ git:
   # tags if there are more than one tag in the same commit.
   #
   # Default: `-version:refname`
-  tag_sort: -version:creatordate
+  tag_sort: -version:creatordate

ADOPTERS.md

@@ -17,6 +17,7 @@ If you're using Velero and want to add your organization to this list,
 <a href="https://www.replicated.com/" border="0" target="_blank"><img alt="replicated.com" src="site/static/img/adopters/replicated-logo-red.svg" height="50"></a>
 <a href="https://cloudcasa.io/" border="0" target="_blank"><img alt="cloudcasa.io" src="site/static/img/adopters/cloudcasa.svg" height="50"></a>
 <a href="https://azure.microsoft.com/" border="0" target="_blank"><img alt="azure.com" src="site/static/img/adopters/azure.svg" height="50"></a>
+<a href="https://www.broadcom.com/" border="0" target="_blank"><img alt="broadcom.com" src="site/static/img/adopters/broadcom.svg" height="50"></a>
 ## Success Stories
 
 Below is a list of adopters of Velero in **production environments** that have
@@ -68,6 +69,9 @@ Replicated uses the Velero open source project to enable snapshots in [KOTS][101
 **[Microsoft Azure][105]**<br>
 [Azure Backup for AKS][106] is an Azure native, Kubernetes aware, Enterprise ready backup for containerized applications deployed on Azure Kubernetes Service (AKS). AKS Backup utilizes Velero to perform backup and restore operations to protect stateful applications in AKS clusters.<br>
 
+**[Broadcom][107]**<br>
+[VMware Cloud Foundation][108] (VCF) offers built-in [vSphere Kubernetes Service][109] (VKS),  a Kubernetes runtime that includes a CNCF certified Kubernetes distribution, to deploy and manage containerized workloads. VCF empowers platform engineers with native [Kubernetes multi-cluster management][110] capability for managing Kubernetes (K8s) infrastructure at scale. VCF utilizes Velero for Kubernetes data protection enabling platform engineers to back up and restore containerized workloads manifests & persistent volumes, helping to increase the resiliency of stateful applications in VKS cluster.
+
 ## Adding your organization to the list of Velero Adopters
 
 If you are using Velero and would like to be included in the list of `Velero Adopters`, add an SVG version of your logo to the `site/static/img/adopters` directory in this repo and submit a [pull request][3] with your change. Name the image file something that reflects your company (e.g., if your company is called Acme, name the image acme.png). See this for an example [PR][4].
@@ -125,3 +129,8 @@ If you would like to add your logo to a future `Adopters of Velero` section on [
 
 [105]: https://azure.microsoft.com/
 [106]: https://learn.microsoft.com/azure/backup/backup-overview
+
+[107]: https://www.broadcom.com/
+[108]: https://www.vmware.com/products/cloud-infrastructure/vmware-cloud-foundation
+[109]: https://www.vmware.com/products/cloud-infrastructure/vsphere-kubernetes-service
+[110]: https://blogs.vmware.com/cloud-foundation/2025/09/29/empowering-platform-engineers-with-native-kubernetes-multi-cluster-management-in-vmware-cloud-foundation/

CHANGELOG.md

@@ -1,7 +1,8 @@
 ## Current release:
-  * [CHANGELOG-1.14.md][24]
+  * [CHANGELOG-1.15.md][25]
 
 ## Older releases:
+  * [CHANGELOG-1.14.md][24]
   * [CHANGELOG-1.13.md][23]
   * [CHANGELOG-1.12.md][22]
   * [CHANGELOG-1.11.md][21]
@@ -27,6 +28,7 @@
   * [CHANGELOG-0.3.md][1]
 
 
+[25]: https://github.com/vmware-tanzu/velero/blob/main/changelogs/CHANGELOG-1.15.md
 [24]: https://github.com/vmware-tanzu/velero/blob/main/changelogs/CHANGELOG-1.14.md
 [23]: https://github.com/vmware-tanzu/velero/blob/main/changelogs/CHANGELOG-1.13.md
 [22]: https://github.com/vmware-tanzu/velero/blob/main/changelogs/CHANGELOG-1.12.md

CODE_OF_CONDUCT.md

@@ -5,7 +5,7 @@
 We as members, contributors, and leaders pledge to make participation in the Velero project and our
 community a harassment-free experience for everyone, regardless of age, body
 size, visible or invisible disability, ethnicity, sex characteristics, gender
-identity and expression, level of experience, education, socio-economic status,
+identity and expression, level of experience, education, socioeconomic status,
 nationality, personal appearance, race, religion, or sexual identity
 and orientation.
 

Dockerfile

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 # Velero binary build section
-FROM --platform=$BUILDPLATFORM golang:1.22-bookworm as velero-builder
+FROM --platform=$BUILDPLATFORM golang:1.25-trixie AS velero-builder
 
 ARG GOPROXY
 ARG BIN
@@ -42,33 +42,12 @@ RUN mkdir -p /output/usr/bin && \
     export GOARM=$( echo "${GOARM}" | cut -c2-) && \
     go build -o /output/${BIN} \
     -ldflags "${LDFLAGS}" ${PKG}/cmd/${BIN} && \
+    go build -o /output/velero-restore-helper \
+    -ldflags "${LDFLAGS}" ${PKG}/cmd/velero-restore-helper && \
     go build -o /output/velero-helper \
     -ldflags "${LDFLAGS}" ${PKG}/cmd/velero-helper && \
     go clean -modcache -cache
 
-# Restic binary build section
-FROM --platform=$BUILDPLATFORM golang:1.22-bookworm as restic-builder
-
-ARG BIN
-ARG TARGETOS
-ARG TARGETARCH
-ARG TARGETVARIANT
-ARG RESTIC_VERSION
-
-ENV CGO_ENABLED=0 \
-    GO111MODULE=on \
-    GOPROXY=${GOPROXY} \
-    GOOS=${TARGETOS} \
-    GOARCH=${TARGETARCH} \
-    GOARM=${TARGETVARIANT}
-
-COPY . /go/src/github.com/vmware-tanzu/velero
-
-RUN mkdir -p /output/usr/bin && \
-    export GOARM=$(echo "${GOARM}" | cut -c2-) && \
-    /go/src/github.com/vmware-tanzu/velero/hack/build-restic.sh && \
-    go clean -modcache -cache
-
 # Velero image packing section
 FROM paketobuildpacks/run-jammy-tiny:latest
 
@@ -76,7 +55,4 @@ LABEL maintainer="Xun Jiang <jxun@vmware.com>"
 
 COPY --from=velero-builder /output /
 
-COPY --from=restic-builder /output /
-
 USER cnb:cnb
-

Dockerfile-Windows

@@ -0,0 +1,57 @@
+# Copyright the Velero contributors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+ARG OS_VERSION=1809
+
+# Velero binary build section
+FROM --platform=$BUILDPLATFORM golang:1.25-trixie AS velero-builder
+
+ARG GOPROXY
+ARG BIN
+ARG PKG
+ARG VERSION
+ARG REGISTRY
+ARG GIT_SHA
+ARG GIT_TREE_STATE
+ARG TARGETOS
+ARG TARGETARCH
+ARG TARGETVARIANT
+
+ENV CGO_ENABLED=0 \
+    GO111MODULE=on \
+    GOPROXY=${GOPROXY} \
+    GOOS=${TARGETOS} \
+    GOARCH=${TARGETARCH} \
+    GOARM=${TARGETVARIANT} \
+    LDFLAGS="-X ${PKG}/pkg/buildinfo.Version=${VERSION} -X ${PKG}/pkg/buildinfo.GitSHA=${GIT_SHA} -X ${PKG}/pkg/buildinfo.GitTreeState=${GIT_TREE_STATE} -X ${PKG}/pkg/buildinfo.ImageRegistry=${REGISTRY}"
+
+WORKDIR /go/src/github.com/vmware-tanzu/velero
+
+COPY . /go/src/github.com/vmware-tanzu/velero
+
+RUN mkdir -p /output/usr/bin && \
+    export GOARM=$( echo "${GOARM}" | cut -c2-) && \
+    go build -o /output/${BIN}.exe \
+    -ldflags "${LDFLAGS}" ${PKG}/cmd/${BIN} && \
+    go build -o /output/velero-restore-helper.exe \
+    -ldflags "${LDFLAGS}" ${PKG}/cmd/velero-restore-helper && \    
+    go build -o /output/velero-helper.exe \
+    -ldflags "${LDFLAGS}" ${PKG}/cmd/velero-helper && \
+    go clean -modcache -cache
+
+# Velero image packing section
+FROM mcr.microsoft.com/windows/nanoserver:${OS_VERSION}
+COPY --from=velero-builder /output /
+
+USER ContainerUser

GOVERNANCE.md

@@ -107,6 +107,29 @@ Lazy consensus does _not_ apply to the process of:
 
 * Removal of maintainers from Velero
 
+## Deprecation Policy
+
+### Deprecation Process
+
+Any contributor may introduce a request to deprecate a feature or an option of a feature by opening a feature request issue in the vmware-tanzu/velero GitHub project. The issue should describe why the feature is no longer needed or has become detrimental to Velero, as well as whether and how it has been superseded. The submitter should give as much detail as possible.
+
+Once the issue is filed, a one-month discussion period begins. Discussions take place within the issue itself as well as in the community meetings. The person who opens the issue, or a maintainer, should add the date and time marking the end of the discussion period in a comment on the issue as soon as possible after it is opened. A decision on the issue needs to be made within this one-month period.
+
+The feature will be deprecated by a supermajority vote of 50% plus one of the project maintainers at the time of the vote tallying, which is 72 hours after the end of the community meeting that is the end of the comment period. (Maintainers are permitted to vote in advance of the deadline, but should hold their votes until as close as possible to hear all possible discussion.) Votes will be tallied in comments on the issue. 
+
+Non-maintainers may add non-binding votes in comments to the issue as well; these are opinions to be taken into consideration by maintainers, but they do not count as votes. 
+
+If the vote passes, the deprecation window takes effect in the subsequent release, and the removal follows the schedule. 
+
+### Schedule
+If depreciation proposal passes by supermajority votes, the feature is deprecated in the next minor release and the feature can be removed completely after two minor version or equivalent major version e.g., if feature gets deprecated in Nth minor version, then feature can be removed after N+2 minor version or its equivalent if the major version number changes.
+
+### Deprecation Window
+
+The deprecation window is the period from the release in which the deprecation takes effect through the release in which the feature is removed. During this period, only critical security vulnerabilities and catastrophic bugs should be fixed.
+
+**Note:** If a backup relies on a deprecated feature, then backups made with the last Velero release before this feature is removed must still be restorable in version `n+2`. For instance, something like restic feature support, that might mean that restic is removed from the list of supported uploader types in version `n` but the underlying implementation required to restore from a restic backup won't be removed until release `n+2`.
+
 ## Updating Governance
 
 All substantive changes in Governance require a supermajority agreement by all maintainers.

MAINTAINERS.md

@@ -7,13 +7,13 @@
 | Maintainer          | GitHub ID                                                     | Affiliation                                      |
 |---------------------|---------------------------------------------------------------|--------------------------------------------------|
 | Scott Seago         | [sseago](https://github.com/sseago)                           | [OpenShift](https://github.com/openshift)        |
-| Daniel Jiang        | [reasonerjt](https://github.com/reasonerjt)                   | [VMware](https://www.github.com/vmware/)         |
-| Wenkai Yin          | [ywk253100](https://github.com/ywk253100)                     | [VMware](https://www.github.com/vmware/)         |
-| Xun Jiang           | [blackpiglet](https://github.com/blackpiglet)                 | [VMware](https://www.github.com/vmware/)         |
-| Ming Qiu            | [qiuming-best](https://github.com/qiuming-best)               | [VMware](https://www.github.com/vmware/)         |
+| Daniel Jiang        | [reasonerjt](https://github.com/reasonerjt)                   | Broadcom                                         |
+| Wenkai Yin          | [ywk253100](https://github.com/ywk253100)                     | Broadcom                                         |
+| Xun Jiang           | [blackpiglet](https://github.com/blackpiglet)                 | Broadcom                                         |
 | Shubham Pampattiwar | [shubham-pampattiwar](https://github.com/shubham-pampattiwar) | [OpenShift](https://github.com/openshift)        |
-| Yonghui Li          | [Lyndon-Li](https://github.com/Lyndon-Li)                     | [VMware](https://www.github.com/vmware/)         |
+| Yonghui Li          | [Lyndon-Li](https://github.com/Lyndon-Li)                     | Broadcom                                         |
 | Anshul Ahuja        | [anshulahuja98](https://github.com/anshulahuja98)             | [Microsoft Azure](https://www.github.com/azure/) |
+| Tiger Kaovilai      | [kaovilai](https://github.com/kaovilai)                       | [OpenShift](https://github.com/openshift)        |
 
 ## Emeritus Maintainers
 * Adnan Abdulhussein ([prydonius](https://github.com/prydonius))
@@ -26,14 +26,4 @@
 * Bridget McErlean ([zubron](https://github.com/zubron))
 * JenTing Hsiao ([jenting](https://github.com/jenting))
 * Dave Smith-Uchida ([dsu-igeek](https://github.com/dsu-igeek))
-  
-## Velero Contributors & Stakeholders
-
-| Feature Area           |                                         Lead                                         |
-|------------------------|:------------------------------------------------------------------------------------:|
-| Technical Lead         |               Daniel Jiang [reasonerjt](https://github.com/reasonerjt)               |
-| Kubernetes CSI Liaison |                                                                                      |
-| Deployment             |                                                                                      |
-| Community Management   |            Orlin Vasilev [OrlinVasilev](https://github.com/OrlinVasilev)             |
-| Product Management     | Pradeep Kumar Chaturvedi [pradeepkchaturvedi](https://github.com/pradeepkchaturvedi) |
-
+* Ming Qiu ([qiuming-best](https://github.com/qiuming-best))

Makefile

@@ -22,15 +22,26 @@ PKG := github.com/vmware-tanzu/velero
 
 # Where to push the docker image.
 REGISTRY ?= velero
-GCR_REGISTRY ?= gcr.io/velero-gcp
+# In order to push images to an insecure registry, follow the two steps:
+#   1. Set "INSECURE_REGISTRY=true" 
+#   2. Provide your own buildx builder instance by setting "BUILDX_INSTANCE=your-own-builder-instance"
+#      The builder can be created with the following command:
+#        cat << EOF > buildkitd.toml
+#        [registry."insecure-registry-ip:port"]
+#        http = true
+#        insecure = true
+#        EOF
+#        docker buildx create --name=velero-builder --driver=docker-container --bootstrap --use --config ./buildkitd.toml
+#      Refer to https://github.com/docker/buildx/issues/1370#issuecomment-1288516840 for more details
+INSECURE_REGISTRY ?= false
 
 # Image name
 IMAGE ?= $(REGISTRY)/$(BIN)
-GCR_IMAGE ?= $(GCR_REGISTRY)/$(BIN)
 
 # We allow the Dockerfile to be configurable to enable the use of custom Dockerfiles
 # that pull base images from different registries.
 VELERO_DOCKERFILE ?= Dockerfile
+VELERO_DOCKERFILE_WINDOWS ?= Dockerfile-Windows
 BUILDER_IMAGE_DOCKERFILE ?= hack/build-image/Dockerfile
 
 # Calculate the realpath of the build-image Dockerfile as we `cd` into the hack/build
@@ -54,7 +65,7 @@ endif
 BUILDER_IMAGE := $(REGISTRY)/build-image:$(BUILDER_IMAGE_TAG)
 BUILDER_IMAGE_CACHED := $(shell docker images -q ${BUILDER_IMAGE} 2>/dev/null )
 
-HUGO_IMAGE := hugo-builder
+HUGO_IMAGE := ghcr.io/gohugoio/hugo
 
 # Which architecture to build - see $(ALL_ARCH) for options.
 # if the 'local' rule is being run, detect the ARCH from 'go env'
@@ -68,23 +79,21 @@ TAG_LATEST ?= false
 
 ifeq ($(TAG_LATEST), true)
 	IMAGE_TAGS ?= $(IMAGE):$(VERSION) $(IMAGE):latest
-	GCR_IMAGE_TAGS ?= $(GCR_IMAGE):$(VERSION) $(GCR_IMAGE):latest
 else
 	IMAGE_TAGS ?= $(IMAGE):$(VERSION)
-	GCR_IMAGE_TAGS ?= $(GCR_IMAGE):$(VERSION)
 endif
 
-# check buildx is enabled
+# check buildx is enabled only if docker is in path
 # macOS/Windows docker cli without Docker Desktop license: https://github.com/abiosoft/colima
 # To add buildx to docker cli: https://github.com/abiosoft/colima/discussions/273#discussioncomment-2684502
-ifeq ($(shell docker buildx inspect 2>/dev/null | awk '/Status/ { print $$2 }'), running)
+ifeq ($(shell which docker 2>/dev/null 1>&2 && docker buildx inspect 2>/dev/null | awk '/Status/ { print $$2 }'), running)
 	BUILDX_ENABLED ?= true
 # if emulated docker cli from podman, assume enabled
 # emulated docker cli from podman: https://podman-desktop.io/docs/migrating-from-docker/emulating-docker-cli-with-podman
 # podman known issues:
 # - on remote podman, such as on macOS,
 #   --output issue: https://github.com/containers/podman/issues/15922
-else ifeq ($(shell cat $(shell which docker) | grep -c "exec podman"), 1)
+else ifeq ($(shell which docker 2>/dev/null 1>&2 && cat $(shell which docker) | grep -c "exec podman"), 1)
 	BUILDX_ENABLED ?= true
 else
 	BUILDX_ENABLED ?= false
@@ -94,13 +103,30 @@ define BUILDX_ERROR
 buildx not enabled, refusing to run this recipe
 see: https://velero.io/docs/main/build-from-source/#making-images-and-updating-velero for more info
 endef
+# comma cannot be escaped and can only be used in Make function arguments by putting into variable
+comma=,
 
-# The version of restic binary to be downloaded
-RESTIC_VERSION ?= 0.15.0
+CLI_PLATFORMS ?= linux-amd64 linux-arm linux-arm64 darwin-amd64 darwin-arm64 windows-amd64 linux-ppc64le linux-s390x
+BUILD_OUTPUT_TYPE ?= docker
+BUILD_OS ?= linux
+BUILD_ARCH ?= amd64
+BUILD_WINDOWS_VERSION ?= ltsc2022
 
-CLI_PLATFORMS ?= linux-amd64 linux-arm linux-arm64 darwin-amd64 darwin-arm64 windows-amd64 linux-ppc64le
-BUILDX_PLATFORMS ?= $(subst -,/,$(ARCH))
-BUILDX_OUTPUT_TYPE ?= docker
+ifeq ($(BUILD_OUTPUT_TYPE), docker)
+	ALL_OS = linux
+	ALL_ARCH.linux = $(word 2, $(subst -, ,$(shell go env GOOS)-$(shell go env GOARCH)))
+else
+	ALL_OS = $(subst $(comma), ,$(BUILD_OS))
+	ALL_ARCH.linux = $(subst $(comma), ,$(BUILD_ARCH))
+endif
+
+ALL_ARCH.windows = $(if $(filter windows,$(ALL_OS)),amd64,)
+ALL_OSVERSIONS.windows = $(if $(filter windows,$(ALL_OS)),$(BUILD_WINDOWS_VERSION),)
+ALL_OS_ARCH.linux =  $(foreach os, $(filter linux,$(ALL_OS)), $(foreach arch, ${ALL_ARCH.linux}, ${os}-$(arch)))
+ALL_OS_ARCH.windows = $(foreach os, $(filter windows,$(ALL_OS)), $(foreach arch, $(ALL_ARCH.windows), $(foreach osversion, ${ALL_OSVERSIONS.windows}, ${os}-${osversion}-${arch})))
+ALL_OS_ARCH = $(ALL_OS_ARCH.linux)$(ALL_OS_ARCH.windows)
+
+ALL_IMAGE_TAGS = $(IMAGE_TAGS)
 
 # set git sha and tree state
 GIT_SHA = $(shell git rev-parse HEAD)
@@ -124,17 +150,14 @@ GOBIN=$$(pwd)/.go/bin
 # If you want to build all containers, see the 'all-containers' rule.
 all:
 	@$(MAKE) build
-	@$(MAKE) build BIN=velero-restore-helper
 
 build-%:
 	@$(MAKE) --no-print-directory ARCH=$* build
-	@$(MAKE) --no-print-directory ARCH=$* build BIN=velero-restore-helper
 
 all-build: $(addprefix build-, $(CLI_PLATFORMS))
 
 all-containers:
 	@$(MAKE) --no-print-directory container
-	@$(MAKE) --no-print-directory container BIN=velero-restore-helper
 
 local: build-dirs
 # Add DEBUG=1 to enable debug locally
@@ -196,11 +219,38 @@ container:
 ifneq ($(BUILDX_ENABLED), true)
 	$(error $(BUILDX_ERROR))
 endif
+
+ifeq ($(BUILDX_INSTANCE),)
+	@echo creating a buildx instance
+	-docker buildx rm velero-builder || true
+	@docker buildx create --use --name=velero-builder
+else
+	@echo using a specified buildx instance $(BUILDX_INSTANCE)
+	@docker buildx use $(BUILDX_INSTANCE)
+endif
+
+	@mkdir -p _output
+
+	@for osarch in $(ALL_OS_ARCH); do \
+		$(MAKE) container-$${osarch}; \
+	done
+
+ifeq ($(BUILD_OUTPUT_TYPE), registry)
+	@for tag in $(ALL_IMAGE_TAGS); do \
+		IMAGE_TAG=$${tag} $(MAKE) push-manifest; \
+	done
+endif
+
+container-linux-%:
+	@BUILDX_ARCH=$* $(MAKE) container-linux
+
+container-linux:
+	@echo "building container: $(IMAGE):$(VERSION)-linux-$(BUILDX_ARCH)"
+
 	@docker buildx build --pull \
-	--output=type=$(BUILDX_OUTPUT_TYPE) \
-	--platform $(BUILDX_PLATFORMS) \
-	$(addprefix -t , $(IMAGE_TAGS)) \
-	$(addprefix -t , $(GCR_IMAGE_TAGS)) \
+	--output="type=$(BUILD_OUTPUT_TYPE)$(if $(findstring tar, $(BUILD_OUTPUT_TYPE)),$(comma)dest=_output/$(BIN)-$(VERSION)-linux-$(BUILDX_ARCH).tar,)" \
+	--platform="linux/$(BUILDX_ARCH)" \
+	$(addprefix -t , $(addsuffix "-linux-$(BUILDX_ARCH)",$(ALL_IMAGE_TAGS))) \
 	--build-arg=GOPROXY=$(GOPROXY) \
 	--build-arg=PKG=$(PKG) \
 	--build-arg=BIN=$(BIN) \
@@ -208,15 +258,54 @@ endif
 	--build-arg=GIT_SHA=$(GIT_SHA) \
 	--build-arg=GIT_TREE_STATE=$(GIT_TREE_STATE) \
 	--build-arg=REGISTRY=$(REGISTRY) \
-	--build-arg=RESTIC_VERSION=$(RESTIC_VERSION) \
+	--provenance=false \
+	--sbom=false \
 	-f $(VELERO_DOCKERFILE) .
-	@echo "container: $(IMAGE):$(VERSION)"
-ifeq ($(BUILDX_OUTPUT_TYPE)_$(REGISTRY), registry_velero)
-	docker pull $(IMAGE):$(VERSION)
-	rm -f $(BIN)-$(VERSION).tar
-	docker save $(IMAGE):$(VERSION) -o $(BIN)-$(VERSION).tar
-	gzip -f $(BIN)-$(VERSION).tar
-endif
+
+	@echo "built container: $(IMAGE):$(VERSION)-linux-$(BUILDX_ARCH)"
+
+container-windows-%:
+	@BUILDX_OSVERSION=$(firstword $(subst -, ,$*)) BUILDX_ARCH=$(lastword $(subst -, ,$*)) $(MAKE) container-windows
+
+container-windows:
+	@echo "building container: $(IMAGE):$(VERSION)-windows-$(BUILDX_OSVERSION)-$(BUILDX_ARCH)"
+
+	@docker buildx build --pull \
+	--output="type=$(BUILD_OUTPUT_TYPE)$(if $(findstring tar, $(BUILD_OUTPUT_TYPE)),$(comma)dest=_output/$(BIN)-$(VERSION)-windows-$(BUILDX_OSVERSION)-$(BUILDX_ARCH).tar,)" \
+	--platform="windows/$(BUILDX_ARCH)" \
+	$(addprefix -t , $(addsuffix "-windows-$(BUILDX_OSVERSION)-$(BUILDX_ARCH)",$(ALL_IMAGE_TAGS))) \
+	--build-arg=GOPROXY=$(GOPROXY) \
+	--build-arg=PKG=$(PKG) \
+	--build-arg=BIN=$(BIN) \
+	--build-arg=VERSION=$(VERSION) \
+	--build-arg=OS_VERSION=$(BUILDX_OSVERSION) \
+	--build-arg=GIT_SHA=$(GIT_SHA) \
+    --build-arg=GIT_TREE_STATE=$(GIT_TREE_STATE) \
+	--build-arg=REGISTRY=$(REGISTRY) \
+	--provenance=false \
+	--sbom=false \
+	-f $(VELERO_DOCKERFILE_WINDOWS) .
+
+	@echo "built container: $(IMAGE):$(VERSION)-windows-$(BUILDX_OSVERSION)-$(BUILDX_ARCH)"
+
+push-manifest:
+	@echo "building manifest: $(IMAGE_TAG) for $(foreach osarch, $(ALL_OS_ARCH), $(IMAGE_TAG)-${osarch})"
+	@docker manifest create --amend --insecure=$(INSECURE_REGISTRY) $(IMAGE_TAG) $(foreach osarch, $(ALL_OS_ARCH), $(IMAGE_TAG)-${osarch})
+
+	@set -x; \
+	for arch in $(ALL_ARCH.windows); do \
+		for osversion in $(ALL_OSVERSIONS.windows); do \
+			BASEIMAGE=mcr.microsoft.com/windows/nanoserver:$${osversion}; \
+			full_version=`docker manifest inspect --insecure=$(INSECURE_REGISTRY) $${BASEIMAGE} | jq -r '.manifests[0].platform["os.version"]'`; \
+			docker manifest annotate --os windows --arch $${arch} --os-version $${full_version} $(IMAGE_TAG) $(IMAGE_TAG)-windows-$${osversion}-$${arch}; \
+		done; \
+	done
+
+	@echo "pushing manifest $(IMAGE_TAG)"
+	@docker manifest push --purge --insecure=$(INSECURE_REGISTRY) $(IMAGE_TAG)
+
+	@echo "pushed manifest $(IMAGE_TAG):"
+	@docker manifest inspect --insecure=$(INSECURE_REGISTRY) $(IMAGE_TAG)
 
 SKIP_TESTS ?=
 test: build-dirs
@@ -359,7 +448,7 @@ release:
 serve-docs: build-image-hugo
 	docker run \
 	--rm \
-	-v "$$(pwd)/site:/srv/hugo" \
+	-v "$$(pwd)/site:/project" \
 	-it -p 1313:1313 \
 	$(HUGO_IMAGE) \
 	server --bind=0.0.0.0 --enableGitInfo=false
@@ -377,4 +466,22 @@ test-perf: local
 	$(MAKE) -e VERSION=$(VERSION) -C test/ run-perf
 
 go-generate:
-	go generate ./pkg/...
+	go generate ./pkg/...
+
+# requires an authenticated gh cli
+# gh: https://cli.github.com/
+# First create a PR
+# gh pr create --title 'Title name' --body 'PR body'
+# by default uses PR title as changelog body but can be overwritten like so
+# make new-changelog CHANGELOG_BODY="Changes you have made"
+new-changelog: GH_LOGIN ?= $(shell gh pr view --json author --jq .author.login 2> /dev/null)
+new-changelog: GH_PR_NUMBER ?= $(shell gh pr view --json number --jq .number 2> /dev/null)
+new-changelog: CHANGELOG_BODY ?= '$(shell gh pr view --json title --jq .title)'
+new-changelog:
+	@if [ "$(GH_LOGIN)" = "" ]; then \
+		echo "branch does not have PR or cli not logged in, try 'gh auth login' or 'gh pr create'"; \
+		exit 1; \
+	fi
+	@mkdir -p ./changelogs/unreleased/ && \
+	echo $(CHANGELOG_BODY) > ./changelogs/unreleased/$(GH_PR_NUMBER)-$(GH_LOGIN) && \
+	echo \"$(CHANGELOG_BODY)\" added to "./changelogs/unreleased/$(GH_PR_NUMBER)-$(GH_LOGIN)"

README.md

@@ -42,11 +42,11 @@ The following is a list of the supported Kubernetes versions for each Velero ver
 
 | Velero version | Expected Kubernetes version compatibility | Tested on Kubernetes version        |
 |----------------|-------------------------------------------|-------------------------------------|
+| 1.18           | 1.18-latest                               | 1.33.7, 1.34.1, and 1.35.0          |
+| 1.17           | 1.18-latest                               | 1.31.7, 1.32.3, 1.33.1, and 1.34.0  |
+| 1.16           | 1.18-latest                               | 1.31.4, 1.32.3, and 1.33.0          |
+| 1.15           | 1.18-latest                               | 1.28.8, 1.29.8, 1.30.4 and 1.31.1   |
 | 1.14           | 1.18-latest                               | 1.27.9, 1.28.9, and 1.29.4          |
-| 1.13           | 1.18-latest                               | 1.26.5, 1.27.3, 1.27.8, and 1.28.3  |
-| 1.12           | 1.18-latest                               | 1.25.7, 1.26.5, 1.26.7, and 1.27.3  |
-| 1.11           | 1.18-latest                               | 1.23.10, 1.24.9, 1.25.5, and 1.26.1 |
-| 1.10           | 1.18-latest                               | 1.22.5, 1.23.8, 1.24.6 and 1.25.1   |
 
 Velero supports IPv4, IPv6, and dual stack environments. Support for this was tested against Velero v1.8.
 

SECURITY.md

@@ -12,13 +12,13 @@ The Velero project maintains the following [governance document](https://github.
 
 Security is of the highest importance and all security vulnerabilities or suspected security vulnerabilities should be reported to Velero privately, to minimize attacks against current users of Velero before they are fixed. Vulnerabilities will be investigated and patched on the next patch (or minor) release as soon as possible. This information could be kept entirely internal to the project.  
 
-If you know of a publicly disclosed security vulnerability for Velero, please **IMMEDIATELY** contact the VMware Security Team (security@vmware.com).
+If you know of a publicly disclosed security vulnerability for Velero, please **IMMEDIATELY** contact the Security Team (velero-security.pdl@broadcom.com).
 
  
 
 **IMPORTANT: Do not file public issues on GitHub for security vulnerabilities**
 
-To report a vulnerability or a security-related issue, please contact the VMware email address with the details of the vulnerability. The email will be fielded by the VMware Security Team and then shared with the Velero maintainers who have committer and release permissions. Emails will be addressed within 3 business days, including a detailed plan to investigate the issue and any potential workarounds to perform in the meantime. Do not report non-security-impacting bugs through this channel. Use [GitHub issues](https://github.com/vmware-tanzu/velero/issues/new/choose) instead.
+To report a vulnerability or a security-related issue, please contact the email address with the details of the vulnerability. The email will be fielded by the Security Team and then shared with the Velero maintainers who have committer and release permissions. Emails will be addressed within 3 business days, including a detailed plan to investigate the issue and any potential workarounds to perform in the meantime. Do not report non-security-impacting bugs through this channel. Use [GitHub issues](https://github.com/vmware-tanzu/velero/issues/new/choose) instead.
 
 
 ## Proposed Email Content
@@ -29,7 +29,7 @@ Provide a descriptive subject line and in the body of the email include the foll
 
 *   Basic identity information, such as your name and your affiliation or company.
 *   Detailed steps to reproduce the vulnerability  (POC scripts, screenshots, and logs are all helpful to us).
-*   Description of the effects of the vulnerability on Velero and the related hardware and software configurations, so that the VMware Security Team can reproduce it.
+*   Description of the effects of the vulnerability on Velero and the related hardware and software configurations, so that the Security Team can reproduce it.
 *   How the vulnerability affects Velero usage and an estimation of the attack surface, if there is one.
 *   List other projects or dependencies that were used in conjunction with Velero to produce the vulnerability.
 
@@ -49,7 +49,7 @@ Provide a descriptive subject line and in the body of the email include the foll
 
 ## Patch, Release, and Disclosure
 
-The VMware Security Team will respond to vulnerability reports as follows:
+The Security Team will respond to vulnerability reports as follows:
 
  
 
@@ -62,7 +62,7 @@ The VMware Security Team will respond to vulnerability reports as follows:
 5. The Security Team will also create a [CVSS](https://www.first.org/cvss/specification-document) using the [CVSS Calculator](https://www.first.org/cvss/calculator/3.0). The Security Team makes the final call on the calculated CVSS; it is better to move quickly than making the CVSS perfect. Issues may also be reported to [Mitre](https://cve.mitre.org/) using this [scoring calculator](https://nvd.nist.gov/vuln-metrics/cvss/v3-calculator). The CVE will initially be set to private.
 6. The Security Team will work on fixing the vulnerability and perform internal testing before preparing to roll out the fix.
 7. The Security Team will provide early disclosure of the vulnerability by emailing the [Velero Distributors](https://groups.google.com/u/1/g/projectvelero-distributors) mailing list. Distributors can initially plan for the vulnerability patch ahead of the fix, and later can test the fix and provide feedback to the Velero team. See the section **Early Disclosure to Velero Distributors List** for details about how to join this mailing list. 
-8. A public disclosure date is negotiated by the VMware SecurityTeam, the bug submitter, and the distributors list. We prefer to fully disclose the bug as soon as possible once a user mitigation or patch is available. It is reasonable to delay disclosure when the bug or the fix is not yet fully understood, the solution is not well-tested, or for distributor coordination. The timeframe for disclosure is from immediate (especially if it’s already publicly known) to a few weeks. For a critical vulnerability with a straightforward mitigation, we expect the report date for the public disclosure date to be on the order of 14 business days. The VMware Security Team holds the final say when setting a public disclosure date.
+8. A public disclosure date is negotiated by the SecurityTeam, the bug submitter, and the distributors list. We prefer to fully disclose the bug as soon as possible once a user mitigation or patch is available. It is reasonable to delay disclosure when the bug or the fix is not yet fully understood, the solution is not well-tested, or for distributor coordination. The timeframe for disclosure is from immediate (especially if it’s already publicly known) to a few weeks. For a critical vulnerability with a straightforward mitigation, we expect the report date for the public disclosure date to be on the order of 14 business days. The Security Team holds the final say when setting a public disclosure date.
 9. Once the fix is confirmed, the Security Team will patch the vulnerability in the next patch or minor release, and backport a patch release into all earlier supported releases. Upon release of the patched version of Velero, we will follow the **Public Disclosure Process**.
 
 
@@ -79,7 +79,7 @@ The Security Team will also publish any mitigating steps users can take until th
 
 
 
-*   Use security@vmware.com to report security concerns to the VMware Security Team, who uses the list to privately discuss security issues and fixes prior to disclosure.
+*   Use velero-security.pdl@broadcom.com to report security concerns to the Security Team, who uses the list to privately discuss security issues and fixes prior to disclosure.
 *   Join the [Velero Distributors](https://groups.google.com/u/1/g/projectvelero-distributors) mailing list for early private information and vulnerability disclosure. Early disclosure may include mitigating steps and additional information on security patch releases. See below for information on how Velero distributors or vendors can apply to join this list.
 
 
@@ -107,11 +107,11 @@ To be eligible to join the [Velero Distributors](https://groups.google.com/u/1/g
 
 ## Embargo Policy
 
-The information that members receive on the Velero Distributors mailing list must not be made public, shared, or even hinted at anywhere beyond those who need to know within your specific team, unless you receive explicit approval to do so from the VMware Security Team. This remains true until the public disclosure date/time agreed upon by the list. Members of the list and others cannot use the information for any reason other than to get the issue fixed for your respective distribution's users.
+The information that members receive on the Velero Distributors mailing list must not be made public, shared, or even hinted at anywhere beyond those who need to know within your specific team, unless you receive explicit approval to do so from the Security Team. This remains true until the public disclosure date/time agreed upon by the list. Members of the list and others cannot use the information for any reason other than to get the issue fixed for your respective distribution's users.
 
 Before you share any information from the list with members of your team who are required to fix the issue, these team members must agree to the same terms, and only be provided with information on a need-to-know basis.
 
-In the unfortunate event that you share information beyond what is permitted by this policy, you must urgently inform the VMware Security Team (security@vmware.com) of exactly what information was leaked and to whom. If you continue to leak information and break the policy outlined here, you will be permanently removed from the list.
+In the unfortunate event that you share information beyond what is permitted by this policy, you must urgently inform the Security Team (velero-security.pdl@broadcom.com) of exactly what information was leaked and to whom. If you continue to leak information and break the policy outlined here, you will be permanently removed from the list.
 
  
 
@@ -123,6 +123,6 @@ Send new membership requests to projectvelero-distributors@googlegroups.com. In
 
 ## Confidentiality, integrity and availability
 
-We consider vulnerabilities leading to the compromise of data confidentiality, elevation of privilege, or integrity to be our highest priority concerns. Availability, in particular in areas relating to DoS and resource exhaustion, is also a serious security concern. The VMware Security Team takes all vulnerabilities, potential vulnerabilities, and suspected vulnerabilities seriously and will investigate them in an urgent and expeditious manner.
+We consider vulnerabilities leading to the compromise of data confidentiality, elevation of privilege, or integrity to be our highest priority concerns. Availability, in particular in areas relating to DoS and resource exhaustion, is also a serious security concern. The Security Team takes all vulnerabilities, potential vulnerabilities, and suspected vulnerabilities seriously and will investigate them in an urgent and expeditious manner.
 
 Note that we do not currently consider the default settings for Velero to be secure-by-default. It is necessary for operators to explicitly configure settings, role based access control, and other resource related features in Velero to provide a hardened Velero environment. We will not act on any security disclosure that relates to a lack of safe defaults. Over time, we will work towards improved safe-by-default configuration, taking into account backwards compatibility.

Tiltfile

@@ -52,7 +52,7 @@ git_sha = str(local("git rev-parse HEAD", quiet = True, echo_off = True)).strip(
 
 tilt_helper_dockerfile_header = """
 # Tilt image
-FROM golang:1.22 as tilt-helper
+FROM golang:1.25 as tilt-helper
 
 # Support live reloading with Tilt
 RUN wget --output-document /restart.sh --quiet https://raw.githubusercontent.com/windmilleng/rerun-process-wrapper/master/restart.sh  && \
@@ -103,11 +103,6 @@ local_resource(
     deps = ["internal", "pkg/cmd"],
 )
 
-local_resource(
-    "restic_binary",
-    cmd = 'cd ' + '.' + ';mkdir -p _tiltbuild/restic; BIN=velero GOOS=linux GOARCH=amd64 GOARM="" RESTIC_VERSION=0.13.1 OUTPUT_DIR=_tiltbuild/restic ./hack/build-restic.sh',
-)
-
 # Note: we need a distro with a bash shell to exec into the Velero container
 tilt_dockerfile_header = """
 FROM ubuntu:22.04 as tilt
@@ -118,7 +113,6 @@ WORKDIR /
 COPY --from=tilt-helper /start.sh .
 COPY --from=tilt-helper /restart.sh .
 COPY velero .
-COPY restic/restic /usr/bin/restic
 """
 
 dockerfile_contents = "\n".join([

changelogs/CHANGELOG-1.14.md

@@ -14,7 +14,7 @@ https://velero.io/docs/v1.14/upgrade-to-1.14/
 
 ### Highlights
 
-#### The maintenance work for kopia backup repositories is run in jobs
+#### The maintenance work for kopia/restic backup repositories is run in jobs
 Since velero started using kopia as the approach for filesystem-level backup/restore, we've noticed an issue when velero connects to the kopia backup repositories and performs maintenance, it sometimes consumes excessive memory that can cause the velero pod to get OOM Killed.  To mitigate this issue, the maintenance work will be moved out of velero pod to a separate kubernetes job, and the user will be able to specify the resource request in "velero install".
 #### Volume Policies are extended to support more actions to handle volumes
 In an earlier release, a flexible volume policy was introduced to skip certain volumes from a backup.  In v1.14 we've made enhancement to this policy to allow the user to set how the volumes should be backed up.  The user will be able to set "fs-backup" or "snapshot" as value of “action" in the policy and velero will backup the volumes accordingly.  This enhancement allows the user to achieve a fine-grained control like "opt-in/out" without having to update the target workload.  For more details please refer to https://velero.io/docs/v1.14/resource-filtering/#supported-volumepolicy-actions
@@ -38,6 +38,7 @@ Besides the service principal with secret(password)-based authentication, Velero
 * CSI plugin has been merged into velero repo in v1.14 release.  It will be installed by default as an internal plugin, and should not be installed via "–plugins " parameter in "velero install" command.
 * The default resource requests and limitations for node agent are removed in v1.14, to make the node agent pods have the QoS class of "BestEffort", more details please refer to #7391
 * There's a change in namespace filtering behavior during backup:  In v1.14, when the includedNamespaces/excludedNamespaces fields are not set and the labelSelector/OrLabelSelectors are set in the backup spec, the backup will only include the namespaces which contain the resources that match the label selectors, while in previous releases all namespaces will be included in the backup with such settings.  More details refer to #7105
+* Patching the PV in the "Finalizing" state may cause the restore to be in "PartiallyFailed" state when the PV is blocked in "Pending" state, while in the previous release the restore may end up being in "Complete" state. For more details refer to #7866
 
 ### All Changes
 * Fix backup log to show error string, not index (#7805, @piny940)

changelogs/CHANGELOG-1.15.md

@@ -0,0 +1,145 @@
+## v1.15
+
+### Download
+https://github.com/vmware-tanzu/velero/releases/tag/v1.15.0
+
+### Container Image
+`velero/velero:v1.15.0`
+
+### Documentation
+https://velero.io/docs/v1.15/
+
+### Upgrading
+https://velero.io/docs/v1.15/upgrade-to-1.15/
+
+### Highlights
+#### Data mover micro service
+Data transfer activities for CSI Snapshot Data Movement are moved from node-agent pods to dedicate backupPods or restorePods. This brings many benefits such as:  
+- This avoids to access volume data through host path, while host path access is privileged and may involve security escalations, which are concerned by users.
+- This enables users to to control resource (i.e., cpu, memory) allocations in a granular manner, e.g., control them per backup/restore of a volume.
+- This enhances the resilience, crash of one data movement activity won't affect others.
+- This prevents unnecessary full backup because of host path changes after workload pods restart.
+- For more information, check the design https://github.com/vmware-tanzu/velero/blob/main/design/Implemented/vgdp-micro-service/vgdp-micro-service.md.
+
+#### Item Block concepts and ItemBlockAction (IBA) plugin
+Item Block concepts are introduced for resource backups to help to achieve multiple thread backups. Specifically, correlated resources are categorized in the same item block and item blocks could be processed concurrently in multiple threads.  
+ItemBlockAction plugin is introduced to help Velero to categorize resources into item blocks. At present, Velero provides built-in IBAs for pods and PVCs and Velero also supports customized IBAs for any resources.  
+In v1.15, Velero doesn't support multiple thread process of item blocks though item block concepts and IBA plugins are fully supported. The multiple thread support will be delivered in future releases.  
+For more information, check the design https://github.com/vmware-tanzu/velero/blob/main/design/backup-performance-improvements.md.  
+
+#### Node selection for repository maintenance job
+Repository maintenance are resource consuming tasks, Velero now allows you to configure the nodes to run repository maintenance jobs, so that you can run repository maintenance jobs in idle nodes or avoid them to run in nodes hosting critical workloads.  
+To support the configuration, a new repository maintenance configuration configMap is introduced.  
+For more information, check the document https://velero.io/docs/v1.15/repository-maintenance/.  
+
+#### Backup PVC read-only configuration
+In 1.15, Velero allows you to configure the data mover backupPods to read-only mount the backupPVCs. In this way, the data mover expose process could be significantly accelerated for some storages (i.e., ceph).  
+To support the configuration, a new backup PVC configuration configMap is introduced.  
+For more information, check the document https://velero.io/docs/v1.15/data-movement-backup-pvc-configuration/.  
+
+#### Backup PVC storage class configuration
+In 1.15, Velero allows you to configure the storageclass used by the data mover backupPods. In this way, the provision of backupPVCs don't need to adhere to the same pattern as workload PVCs, e.g., for a backupPVC, it only needs one replica, whereas, the a workload PVC may have multiple replicas.  
+To support the configuration, the same backup PVC configuration configMap is used.  
+For more information, check the document https://velero.io/docs/v1.15/data-movement-backup-pvc-configuration/.  
+
+#### Backup repository data cache configuration
+The backup repository may need to cache data on the client side during various repository operations, i.e., read, write, maintenance, etc. The cache consumes the root file system space of the pod where the repository access happens.  
+In 1.15, Velero allows you to configure the total size of the cache per repository. In this way, if your pod doesn't have enough space in its root file system, the pod won't be evicted due to running out of ephemeral storage.  
+To support the configuration, a new backup repository configuration configMap is introduced.  
+For more information, check the document https://velero.io/docs/v1.15/backup-repository-configuration/.  
+
+#### Performance improvements
+In 1.15, several performance related issues/enhancements are included, which makes significant performance improvements in specific scenarios:  
+- There was a memory leak of Velero server after plugin calls, now it is fixed, see issue https://github.com/vmware-tanzu/velero/issues/7925
+- The `client-burst/client-qps` parameters are automatically inherited to plugins, so that you can use the same velero server parameters to accelerate the plugin executions when large number of API server calls happen, see issue https://github.com/vmware-tanzu/velero/issues/7806
+- Maintenance of Kopia repository takes huge memory in scenarios that huge number of files have been backed up, Velero 1.15 has included the Kopia upstream enhancement to fix the problem, see issue https://github.com/vmware-tanzu/velero/issues/7510
+
+### Runtime and dependencies
+Golang runtime: v1.22.8  
+kopia: v0.17.0
+
+### Limitations/Known issues
+#### Read-only backup PVC may not work on SELinux environments
+Due to an issue of Kubernetes upstream, if a volume is mounted as read-only in SELinux environments, the read privilege is not granted to any user, as a result, the data mover backup will fail. On the other hand, the backupPVC must be mounted as read-only in order to accelerate the data mover expose process.  
+Therefore, a user option is added in the same backup PVC configuration configMap, once the option is enabled, the backupPod container will run as a super privileged container and disable SELinux access control. If you have concern in this super privileged container or you have configured [pod security admissions](https://kubernetes.io/docs/concepts/security/pod-security-admission/) and don't allow super privileged containers, you will not be able to use this read-only backupPVC feature and lose the benefit to accelerate the data mover expose process.  
+
+### Breaking changes
+#### Deprecation of Restic
+Restic path for fs-backup is in deprecation process starting from 1.15. According to [Velero deprecation policy](https://github.com/vmware-tanzu/velero/blob/v1.15/GOVERNANCE.md#deprecation-policy), for 1.15, if Restic path is used the backup/restore of fs-backup still creates and succeeds, but you will see warnings in below scenarios:  
+- When `--uploader-type=restic` is used in Velero installation
+- When Restic path is used to create backup/restore of fs-backup
+
+#### node-agent configuration name is configurable
+Previously, a fixed name is searched for node-agent configuration configMap. Now in 1.15, Velero allows you to customize the name of the configMap, on the other hand, the name must be specified by node-agent server parameter `node-agent-configmap`.  
+
+#### Repository maintenance job configurations in Velero server parameter are moved to repository maintenance job configuration configMap
+In 1.15, below Velero server parameters for repository maintenance jobs are moved to the repository maintenance job configuration configMap. While for back compatibility reason, the same Velero sever parameters are preserved as is. But the configMap is recommended and the same values in the configMap take preference if they exist in both places:  
+```
+--keep-latest-maintenance-jobs
+--maintenance-job-cpu-request
+--maintenance-job-mem-request
+--maintenance-job-cpu-limit
+--maintenance-job-mem-limit
+```
+
+#### Changing PVC selected-node feature is deprecated
+In 1.15, the [Changing PVC selected-node feature](https://velero.io/docs/v1.15/restore-reference/#changing-pvc-selected-node) enters deprecation process and will be removed in future releases according to [Velero deprecation policy](https://github.com/vmware-tanzu/velero/blob/v1.15/GOVERNANCE.md#deprecation-policy). Usage of this feature for any purpose is not recommended.  
+
+### All Changes
+  * add no-relabeling option to backupPVC configmap (#8288, @sseago)
+  * only set spec.volumes readonly if PVC is readonly for datamover (#8284, @sseago)
+  * Add labels to maintenance job pods (#8256, @shubham-pampattiwar)
+  * Add the Carvel package related resources to the restore priority list (#8228, @ywk253100)
+  * Reduces indirect imports for plugin/framework importers (#8208, @kaovilai)
+  * Add controller name to periodical_enqueue_source. The logger parameter now includes an additional field with the value of reflect.TypeOf(objList).String() and another field with the value of controllerName. (#8198, @kaovilai)
+  * Update Openshift SCC docs link (#8170, @shubham-pampattiwar)
+  * Partially fix issue #8138, add doc for node-agent memory preserve (#8167, @Lyndon-Li)
+  * Pass Velero server command args to the plugins (#8166, @ywk253100)
+  * Fix issue #8155, Merge Kopia upstream commits for critical issue fixes and performance improvements (#8158, @Lyndon-Li)
+  * Implement the Repo maintenance Job configuration. (#8145, @blackpiglet)
+  * Add document for data mover micro service (#8144, @Lyndon-Li)
+  * Fix issue #8134, allow to config resource request/limit for data mover micro service pods (#8143, @Lyndon-Li)
+  * Apply backupPVCConfig to backupPod volume spec (#8141, @shubham-pampattiwar)
+  * Add resource modifier for velero restore describe CLI (#8139, @blackpiglet)
+  * Fix issue #7620, add doc for backup repo config (#8131, @Lyndon-Li)
+  * Modify E2E and perf test report generated directory (#8129, @blackpiglet)
+  * Add docs for backup pvc config support (#8119, @shubham-pampattiwar)
+  * Delete generated k8s client and informer. (#8114, @blackpiglet)
+  * Add support for backup PVC configuration (#8109, @shubham-pampattiwar)
+  * ItemBlock model and phase 1 (single-thread) workflow changes (#8102, @sseago)
+  * Fix issue #8032, make node-agent configMap name configurable (#8097, @Lyndon-Li)
+  * Fix issue #8072, add the warning messages for restic deprecation (#8096, @Lyndon-Li)
+  * Fix issue #7620, add backup repository configuration implementation and support cacheLimit configuration for Kopia repo (#8093, @Lyndon-Li)
+  * Patch dbr's status when error happens (#8086, @reasonerjt)
+  * According to design #7576, after node-agent restarts, if a DU/DD is in InProgress status, re-capture the data mover ms pod and continue the execution (#8085, @Lyndon-Li)
+  * Updates to IBM COS documentation to match current version (#8082, @gjanders)
+  * Data mover micro service DUCR/DDCR controller refactor according to design #7576 (#8074, @Lyndon-Li)
+  * add retries with timeout to existing patch calls that moves a backup/restore from InProgress/Finalizing to a final status phase. (#8068, @kaovilai)
+  * Data mover micro service restore according to design #7576 (#8061, @Lyndon-Li)
+  * Internal ItemBlockAction plugins (#8054, @sseago)
+  * Data mover micro service backup according to design #7576 (#8046, @Lyndon-Li)
+  * Avoid wrapping failed PVB status with empty message. (#8028, @mrnold)
+  * Created new ItemBlockAction (IBA) plugin type (#8026, @sseago)
+  * Make PVPatchMaximumDuration timeout configurable (#8021, @shubham-pampattiwar)
+  * Reuse existing plugin manager for get/put volume info (#8012, @sseago)
+  * Data mover ms watcher according to design #7576 (#7999, @Lyndon-Li)
+  * New data path for data mover ms according to design #7576 (#7988, @Lyndon-Li)
+  * For issue #7700 and #7747, add the design for backup PVC configurations (#7982, @Lyndon-Li)
+  * Only get VolumeSnapshotClass when DataUpload exists. (#7974, @blackpiglet)
+  * Fix issue #7972, sync the backupPVC deletion in expose clean up (#7973, @Lyndon-Li)
+  * Expose the VolumeHelper to third-party plugins. (#7969, @blackpiglet)
+  * Check whether the volume's source is PVC before fetching its PV. (#7967, @blackpiglet)
+  * Check whether the namespaces specified in namespace filter exist. (#7965, @blackpiglet)
+  * Add design for backup repository configurations for issue #7620, #7301 (#7963, @Lyndon-Li)
+  * New data path for data mover ms according to design #7576 (#7955, @Lyndon-Li)
+  * Skip PV patch step in Restoe workflow for WaitForFirstConsumer VolumeBindingMode Pending state PVCs (#7953, @shubham-pampattiwar)
+  * Fix issue #7904, add the deprecation and limitation clarification for change PVC selected-node feature (#7948, @Lyndon-Li)
+  * Expose the VolumeHelper to third-party plugins. (#7944, @blackpiglet)
+  * Don't consider unschedulable pods unrecoverable (#7899, @sseago)
+  * Upgrade to robfig/cron/v3 to support time zone specification. (#7793, @kaovilai)
+  * Add the result in the backup's VolumeInfo. (#7775, @blackpiglet)
+  * Migrate from github.com/golang/protobuf to google.golang.org/protobuf (#7593, @mmorel-35)
+  * Add the design for data mover micro service (#7576, @Lyndon-Li)
+  * Descriptive restore error when restoring into a terminating namespace. (#7424, @kaovilai)
+  * Ignore missing path error in conditional match (#7410, @seanblong)
+  * Propose a deprecation process for velero (#5532, @shubham-pampattiwar)

changelogs/CHANGELOG-1.16.md

@@ -0,0 +1,156 @@
+## v1.16
+
+### Download
+https://github.com/vmware-tanzu/velero/releases/tag/v1.16.0
+
+### Container Image
+`velero/velero:v1.16.0`
+
+### Documentation
+https://velero.io/docs/v1.16/
+
+### Upgrading
+https://velero.io/docs/v1.16/upgrade-to-1.16/
+
+### Highlights
+#### Windows cluster support
+In v1.16, Velero supports to run in Windows clusters and backup/restore Windows workloads, either stateful or stateless:
+ * Hybrid build and all-in-one image: the build process is enhanced to build an all-in-one image for hybrid CPU architecture and hybrid platform. For more information, check the design https://github.com/vmware-tanzu/velero/blob/main/design/multiple-arch-build-with-windows.md
+ * Deployment in Windows clusters: Velero node-agent, data mover pods and maintenance jobs now support to run in both linux and Windows nodes
+ * Data mover backup/restore Windows workloads: Velero built-in data mover supports Windows workloads throughout its full cycle, i.e., discovery, backup, restore, pre/post hook, etc. It automatically identifies Windows workloads and schedules data mover pods to the right group of nodes
+
+Check the epic issue https://github.com/vmware-tanzu/velero/issues/8289 for more information.  
+
+#### Parallel Item Block backup
+v1.16 now supports to back up item blocks in parallel. Specifically, during backup, correlated resources are grouped in item blocks and Velero backup engine creates a thread pool to back up the item blocks in parallel. This significantly improves the backup throughput, especially when there are large scale of resources.  
+Pre/post hooks also belongs to item blocks, so will also run in parallel along with the item blocks.  
+Users are allowed to configure the parallelism through the `--item-block-worker-count` Velero server parameter. If not configured, the default parallelism is 1.  
+
+For more information, check issue https://github.com/vmware-tanzu/velero/issues/8334.  
+
+#### Data mover restore enhancement in scalability
+In previous releases, for each volume of WaitForFirstConsumer mode, data mover restore is only allowed to happen in the node that the volume is attached. This severely degrades the parallelism and the balance of node resource(CPU, memory, network bandwidth) consumption for data mover restore (https://github.com/vmware-tanzu/velero/issues/8044).  
+
+In v1.16, users are allowed to configure data mover restores running and spreading evenly across all nodes in the cluster. The configuration is done through a new flag `ignoreDelayBinding` in node-agent configuration (https://github.com/vmware-tanzu/velero/issues/8242).  
+
+#### Data mover enhancements in observability 
+In 1.16, some observability enhancements are added:
+ * Output various statuses of intermediate objects for failures of data mover backup/restore (https://github.com/vmware-tanzu/velero/issues/8267)
+ * Output the errors when Velero fails to delete intermediate objects during clean up (https://github.com/vmware-tanzu/velero/issues/8125)
+
+The outputs are in the same node-agent log and enabled automatically.  
+
+#### CSI snapshot backup/restore enhancement in usability
+In previous releases, a unnecessary VolumeSnapshotContent object is retained for each backup and synced to other clusters sharing the same backup storage location. And during restore, the retained VolumeSnapshotContent is also restored unnecessarily.  
+
+In 1.16, the retained VolumeSnapshotContent is removed from the backup, so no unnecessary CSI objects are synced or restored.  
+
+For more information, check issue https://github.com/vmware-tanzu/velero/issues/8725.  
+
+#### Backup Repository Maintenance enhancement in resiliency and observability
+In v1.16, some enhancements of backup repository maintenance are added to improve the observability and resiliency:
+ * A new backup repository maintenance history section, called `RecentMaintenance`, is added to the BackupRepository CR. Specifically, for each BackupRepository, including start/completion time, completion status and error message. (https://github.com/vmware-tanzu/velero/issues/7810)
+ * Running maintenance jobs are now recaptured after Velero server restarts. (https://github.com/vmware-tanzu/velero/issues/7753)
+ * The maintenance job will not be launched for readOnly BackupStorageLocation. (https://github.com/vmware-tanzu/velero/issues/8238)
+ * The backup repository will not try to initialize a new repository for readOnly BackupStorageLocation. (https://github.com/vmware-tanzu/velero/issues/8091)
+ * Users now are allowed to configure the intervals of an effective maintenance in the way of `normalGC`, `fastGC` and `eagerGC`, through the `fullMaintenanceInterval` parameter in backupRepository configuration. (https://github.com/vmware-tanzu/velero/issues/8364)
+
+#### Volume Policy enhancement of filtering volumes by PVC labels
+In v1.16, Volume Policy is extended to support filtering volumes by PVC labels. (https://github.com/vmware-tanzu/velero/issues/8256).   
+
+#### Resource Status restore per object
+In v1.16, users are allowed to define whether to restore resource status per object through an annotation `velero.io/restore-status` set on the object. (https://github.com/vmware-tanzu/velero/issues/8204).  
+
+#### Velero Restore Helper binary is merged into Velero image 
+In v1.16, Velero banaries, i.e., velero, velero-helper and velero-restore-helper, are all included into the single Velero image. (https://github.com/vmware-tanzu/velero/issues/8484).  
+
+### Runtime and dependencies
+Golang runtime: 1.23.7  
+kopia: 0.19.0
+
+### Limitations/Known issues
+#### Limitations of Windows support
+  * fs-backup is not supported for Windows workloads and so fs-backup runs only in linux nodes for linux workloads
+  * Backup/restore of NTFS extended attributes/advanced features are not supported, i.e., Security Descriptors, System/Hidden/ReadOnly attributes, Creation Time, NTFS Streams, etc.
+
+### All Changes
+  * Add third party annotation support for maintenance job, so that the declared third party annotations could be added to the maintenance job pods (#8812, @Lyndon-Li)
+  * Fix issue #8803, use deterministic name to create backupRepository (#8808, @Lyndon-Li)
+  * Refactor restoreItem and related functions to differentiate the backup resource name and the restore target resource name. (#8797, @blackpiglet)
+  * ensure that PV is removed before VS is deleted (#8777, @ix-rzi)
+  * host_pods should not be mandatory to node-agent (#8774, @mpryc)
+  * Log doesn't show pv name, but displays %!s(MISSING) instead (#8771, @hu-keyu)
+  * Fix issue #8754, add third party annotation support for data mover (#8770, @Lyndon-Li)
+  * Add docs for volume policy with labels as a criteria (#8759, @shubham-pampattiwar)
+  * Move pvc annotation removal from CSI RIA to regular PVC RIA (#8755, @sseago)
+  * Add doc for maintenance history (#8747, @Lyndon-Li)
+  * Fix issue #8733, add doc for restorePVC (#8737, @Lyndon-Li)
+  * Fix issue #8426, add doc for Windows support (#8736, @Lyndon-Li)
+  * Fix issue #8475, refactor build-from-source doc for hybrid image build (#8729, @Lyndon-Li)
+  * Return directly if no pod volme backup are tracked (#8728, @ywk253100)
+  * Fix issue #8706, for immediate volumes, there is no selected-node annotation on PVC, so deduce the attached node from VolumeAttachment CRs (#8715, @Lyndon-Li)
+  * Add labels as a criteria for volume policy (#8713, @shubham-pampattiwar)
+  * Copy SecurityContext from Containers[0] if present for PVR (#8712, @sseago)
+  * Support pushing images to an insecure registry (#8703, @ywk253100)
+  * Modify golangci configuration to make it work. (#8695, @blackpiglet)
+  * Run backup post hooks inside ItemBlock synchronously (#8694, @ywk253100)
+  * Add docs for object level status restore (#8693, @shubham-pampattiwar)
+  * Clean artifacts generated during CSI B/R. (#8684, @blackpiglet)
+  * Don't run maintenance on the ReadOnly BackupRepositories. (#8681, @blackpiglet)
+  * Fix #8657: WaitGroup panic issue (#8679, @ywk253100)
+  * Fixes issue #8214, validate `--from-schedule` flag in create backup command to prevent empty or whitespace-only values. (#8665, @aj-2000)
+  * Implement parallel ItemBlock processing via backup_controller goroutines (#8659, @sseago)
+  * Clean up leaked CSI snapshot for incomplete backup (#8637, @raesonerjt)
+  * Handle update conflict when restoring the status (#8630, @ywk253100)
+  * Fix issue #8419, support repo maintenance job to run on Windows nodes (#8626, @Lyndon-Li)
+  * Always create DataUpload configmap in restore namespace (#8621, @sseago)
+  * Fix issue #8091, avoid to create new repo when BSL is readonly (#8615, @Lyndon-Li)
+  * Fix issue #8242, distribute dd evenly across nodes (#8611, @Lyndon-Li)
+  * Fix issue #8497, update du/dd progress on completion (#8608, @Lyndon-Li)
+  * Fix issue #8418, add Windows toleration to data mover pods (#8606, @Lyndon-Li)
+  * Check the PVB status via podvolume Backupper rather than calling API server to avoid API server issue (#8603, @ywk253100)
+  * Fix issue #8067, add tmp folder (/tmp for linux, C:\Windows\Temp for Windows) as an alternative of udmrepo's config file location (#8602, @Lyndon-Li)
+  * Data mover restore for Windows (#8594, @Lyndon-Li)
+  * Skip patching the PV in finalization for failed operation (#8591, @reasonerjt)
+  * Fix issue #8579, set event burst to block event broadcaster from filtering events (#8590, @Lyndon-Li)
+  * Configurable Kopia Maintenance Interval. backup-repository-configmap adds an option for configurable`fullMaintenanceInterval` where fastGC (12 hours), and eagerGC (6 hours) allowing for faster removal of deleted velero backups from kopia repo. (#8581, @kaovilai)
+  * Fix issue #7753, recall repo maintenance history on Velero server restart (#8580, @Lyndon-Li)
+  * Clear validation errors when schedule is valid (#8575, @ywk253100)
+  * Merge restore helper image into Velero server image (#8574, @ywk253100)
+  * Don't include excluded items in ItemBlocks (#8572, @sseago)
+  * fs uploader and block uploader support Windows nodes (#8569, @Lyndon-Li)
+  * Fix issue #8418, support data mover backup for Windows nodes (#8555, @Lyndon-Li)
+  * Fix issue #8044, allow users to ignore delay binding the restorePVC of data mover when it is in WaitForFirstConsumer mode (#8550, @Lyndon-Li)
+  * Fix issue #8539, validate uploader types when o.CRDsOnly is set to false only since CRD installation doesn't rely on uploader types (#8538, @Lyndon-Li)
+  * Fix issue #7810, add maintenance history for backupRepository CRs (#8532, @Lyndon-Li)
+  * Make fs-backup work on linux nodes with the new Velero deployment and disable fs-backup if the source/target pod is running in non-linux node (#8424) (#8518, @Lyndon-Li)
+  * Fix issue: backup schedule pause/unpause doesn't work (#8512, @ywk253100)
+  * Fix backup post hook issue #8159 (caused by #7571): always execute backup post hooks after PVBs are handled (#8509, @ywk253100)
+  * Fix issue #8267, enhance the error message when expose fails (#8508, @Lyndon-Li)
+  * Fix issue #8416, #8417, deploy Velero server and node-agent in linux/Windows hybrid env (#8504, @Lyndon-Li)
+  * Design to add label selector as a criteria for volume policy (#8503, @shubham-pampattiwar)
+  * Related to issue #8485, move the acceptedByNode and acceptedTimestamp to Status of DU/DD CRD (#8498, @Lyndon-Li)
+  * Add SecurityContext to restore-helper (#8491, @reasonerjt)
+  * Fix issue #8433, add third party labels to data mover pods when the same labels exist in node-agent pods (#8487, @Lyndon-Li)
+  * Fix issue #8485, add an accepted time so as to count the prepare timeout (#8486, @Lyndon-Li)
+  * Fix issue #8125, log diagnostic info for data mover exposers when expose timeout (#8482, @Lyndon-Li)
+  * Fix issue #8415, implement multi-arch build and Windows build (#8476, @Lyndon-Li)
+  * Pin kopia to 0.18.2 (#8472, @Lyndon-Li)
+  * Add nil check for updating DataUpload VolumeInfo in finalizing phase (#8471, @blackpiglet)
+  * Allowing Object-Level Resource Status Restore (#8464, @shubham-pampattiwar)
+  * For issue #8429. Add the design for multi-arch build and windows build (#8459, @Lyndon-Li)
+  * Upgrade go.mod k8s.io/ go.mod to v0.31.3 and implemented proper logger configuration for both client-go and controller-runtime libraries. This change ensures that logging format and level settings are properly applied throughout the codebase. The update improves logging consistency and control across the Velero system. (#8450, @kaovilai)
+  * Add Design for Allowing Object-Level Resource Status Restore (#8403, @shubham-pampattiwar)
+  * Fix issue #8391, check ErrCancelled from suffix of data mover pod's termination message (#8396, @Lyndon-Li)
+  * Fix issue #8394, don't call closeDataPath in VGDP callbacks, otherwise, the VGDP cleanup will hang (#8395, @Lyndon-Li)
+  * Adding support in velero Resource Policies for filtering PVs based on additional VolumeAttributes properties under CSI PVs (#8383, @mayankagg9722)
+  * Add --item-block-worker-count flag to velero install and server (#8380, @sseago)
+  * Make BackedUpItems thread safe (#8366, @sseago)
+  * Include --annotations flag in backup and restore create commands (#8354, @alromeros)
+  * Use aggregated discovery API to discovery API groups and resources (#8353, @ywk253100)
+  * Copy "envFrom" from Velero server when creating maintenance jobs (#8343, @evhan)
+  * Set hinting region to use for GetBucketRegion() in pkg/repository/config/aws.go (#8297, @kaovilai)
+  * Bump up version of client-go and controller-runtime (#8275, @ywk253100)
+  * fix(pkg/repository/maintenance): don't panic when there's no container statuses (#8271, @mcluseau)
+  * Add Backup warning for inclusion of NS managed by ArgoCD (#8257, @shubham-pampattiwar)
+  * Added tracking for deleted namespace status check in restore flow. (#8233, @sangitaray2021)

changelogs/CHANGELOG-1.17.md

@@ -0,0 +1,143 @@
+## v1.17
+
+### Download
+https://github.com/vmware-tanzu/velero/releases/tag/v1.17.0
+
+### Container Image
+`velero/velero:v1.17.0`
+
+### Documentation
+https://velero.io/docs/v1.17/
+
+### Upgrading
+https://velero.io/docs/v1.17/upgrade-to-1.17/
+
+### Highlights
+#### Modernized fs-backup
+In v1.17, Velero fs-backup is modernized to the micro-service architecture, which brings below benefits:  
+- Many features that were absent to fs-backup are now available, i.e., load concurrency control, cancel, resume on restart, etc.
+- fs-backup is more robust, the running backup/restore could survive from node-agent restart; and the resource allocation is in a more granular manner, the failure of one backup/restore won't impact others.  
+- The resource usage of node-agent is steady, especially, the node-agent pods won't request huge memory and hold it for a long time.  
+
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/vgdp-micro-service-for-fs-backup/vgdp-micro-service-for-fs-backup.md for more details.  
+
+#### fs-backup support Windows cluster
+In v1.17, Velero fs-backup supports to backup/restore Windows workloads. By leveraging the new micro-service architecture for fs-backup, data mover pods could run in Windows nodes and backup/restore Windows volumes. Together with CSI snapshot data movement for Windows which is delivered in 1.16, Velero now supports Windows workload backup/restore in full scenarios.  
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/vgdp-micro-service-for-fs-backup/vgdp-micro-service-for-fs-backup.md for more details.  
+
+#### Volume group snapshot support
+In v1.17, Velero supports [volume group snapshots](https://kubernetes.io/blog/2024/12/18/kubernetes-1-32-volume-group-snapshot-beta/) which is a beta feature in Kubernetes upstream, for both CSI snapshot backup and CSI snapshot data movement. This allows a snapshot to be taken from multiple volumes at the same point-in-time to achieve write order consistency, which is helpful to achieve better data consistency when multiple volumes being backed up are correlated.  
+Check the document https://velero.io/docs/main/volume-group-snapshots/ for more details.  
+
+#### Priority class support
+In v1.17, [Kubernetes priority class](https://kubernetes.io/docs/concepts/scheduling-eviction/pod-priority-preemption/#priorityclass) is supported for all modules across Velero. Specifically, users are allowed to configure priority class to Velero server, node-agent, data mover pods, backup repository maintenance jobs separately.  
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/Implemented/priority-class-name-support_design.md for more details.  
+
+#### Scalability and Resiliency improvements of data movers
+##### Reduce excessive number of data mover pods in Pending state
+In v1.17, Velero allows users to set a `PrepareQueueLength` in the node-agent configuration, data mover pods and volumes out of this number won't be created until data path quota is available, so that excessive number cluster resources won't  be taken unnecessarily, which is particularly helpful for large scale environments. This improvement applies to all kinds of data movements, including fs-backup and CSI snapshot data movement.  
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/node-agent-load-soothing.md for more details.  
+
+##### Enhancement on node-agent restart handling for data movements
+In v1.17, data movements in all phases could survive from node-agent restart and resume themselves; when a data movement gets orphaned in special cases, e.g., cluster node absent, it could also be canceled appropriately after the restart. This improvement applies to all kinds of data movements, including fs-backup and CSI snapshot data movement.  
+Check issue https://github.com/vmware-tanzu/velero/issues/8534 for more details.  
+
+##### CSI snapshot data movement restore node-selection and node-selection by storage class
+In v1.17, CSI snapshot data movement restore acquires the same node-selection capability as backup, that is, users could specify which nodes can/cannot run data mover pods for both backup and restore now. And users are also allowed to configure the node-selection per storage class, which is particularly helpful to the environments where a storage class are not usable by all cluster nodes.  
+Check issue https://github.com/vmware-tanzu/velero/issues/8186 and https://github.com/vmware-tanzu/velero/issues/8223 for more details.  
+
+#### Include/exclude policy support for resource policy
+In v1.17, Velero resource policy supports `includeExcludePolicy` besides the existing `volumePolicy`. This allows users to set include/exclude filters for resources in a resource policy configmap, so that these filters are reusable among multiple backups.  
+Check the document https://velero.io/docs/main/resource-filtering/#creating-resource-policies:~:text=resources%3D%22*%22-,Resource%20policies,-Velero%20provides%20resource for more details.  
+
+### Runtime and dependencies
+Golang runtime: 1.24.6  
+kopia: 0.21.1  
+
+### Limitations/Known issues
+
+### Breaking changes
+#### Deprecation of Restic
+According to [Velero deprecation policy](https://github.com/vmware-tanzu/velero/blob/main/GOVERNANCE.md#deprecation-policy), backup of fs-backup under Restic path is removed in v1.17, so `--uploader-type=restic` is not a valid installation configuration anymore. This means you cannot create a backup under Restic path, but you can still restore from the previous backups under Restic path until v1.19.  
+
+#### Repository maintenance job configurations are removed from Velero server parameter
+Since the repository maintenance job configurations are moved to repository maintenance job configMap, in v1.17 below Velero sever parameters are removed:
+- --keep-latest-maintenance-jobs
+- --maintenance-job-cpu-request
+- --maintenance-job-mem-request
+- --maintenance-job-cpu-limit
+- --maintenance-job-mem-limit
+
+### All Changes
+  * Add ConfigMap parameters validation for install CLI and server start. (#9200, @blackpiglet)
+  * Add priorityclasses to high priority restore list (#9175, @kaovilai)
+  * Introduced context-based logger for backend implementations (Azure, GCS, S3, and Filesystem) (#9168, @priyansh17)
+  * Fix issue #9140, add os=windows:NoSchedule toleration for Windows pods (#9165, @Lyndon-Li)
+  * Remove the repository maintenance job parameters from velero server. (#9147, @blackpiglet)
+  * Add include/exclude policy to resources policy (#9145, @reasonerjt)
+  * Add ConfigMap support for keepLatestMaintenanceJobs with CLI parameter fallback (#9135, @shubham-pampattiwar)
+  * Fix the dd and du's node affinity issue. (#9130, @blackpiglet)
+  * Remove the WaitUntilVSCHandleIsReady from vs BIA. (#9124, @blackpiglet)
+  * Add comprehensive Volume Group Snapshots documentation with workflow diagrams and examples (#9123, @shubham-pampattiwar)
+  * Fix issue #9065, add doc for node-agent prepare queue length (#9118, @Lyndon-Li)
+  * Fix issue #9095, update restore doc for PVC selected-node (#9117, @Lyndon-Li)
+  * Update CSI Snapshot Data Movement doc for issue #8534, #8185 (#9113, @Lyndon-Li)
+  * Fix issue #8986, refactor fs-backup doc after VGDP Micro Service for fs-backup (#9112, @Lyndon-Li)
+  * Return error if timeout when checking server version (#9111, @ywk253100)
+  * Update "Default Volumes to Fs Backup" to "File System Backup (Default)" (#9105, @shubham-pampattiwar)
+  * Fix issue #9077, don't block backup deletion on list VS error (#9100, @Lyndon-Li)
+  * Bump up Kopia to v0.21.1 (#9098, @Lyndon-Li)
+  * Add imagePullSecrets inheritance for VGDP pod and maintenance job. (#9096, @blackpiglet)
+  * Avoid checking the VS and VSC status in the backup finalizing phase. (#9092, @blackpiglet)
+  * Fix issue #9053, Always remove selected-node annotation during PVC restore when no node mapping exists. Breaking change: Previously, the annotation was preserved if the node existed. (#9076, @Lyndon-Li)
+  * Enable parameterized kubelet mount path during node-agent installation (#9074, @longxiucai)
+  * Fix issue #8857, support third party tolerations for data mover pods (#9072, @Lyndon-Li)
+  * Fix issue #8813, remove restic from the valid uploader type (#9069, @Lyndon-Li)
+  * Fix issue #8185, allow users to disable pod volume host path mount for node-agent (#9068, @Lyndon-Li)
+  * Fix #8344, add the design for a mechanism to soothe creation of data mover pods for DataUpload, DataDownload, PodVolumeBackup and PodVolumeRestore (#9067, @Lyndon-Li)
+  * Fix #8344, add a mechanism to soothe creation of data mover pods for DataUpload, DataDownload, PodVolumeBackup and PodVolumeRestore (#9064, @Lyndon-Li)
+  * Add Gauge metric for BSL availability (#9059, @reasonerjt)
+  * Fix missing defaultVolumesToFsBackup flag output in Velero describe backup cmd (#9056, @shubham-pampattiwar)
+  * Allow for proper tracking of multiple hooks per container (#9048, @sseago)
+  * Make the backup repository controller doesn't invalidate the BSL on restart (#9046, @blackpiglet)
+  * Removed username/password credential handling from newConfigCredential as azidentity.UsernamePasswordCredentialOptions is reported as deprecated. (#9041, @priyansh17)
+  * Remove dependency with VolumeSnapshotClass in DataUpload. (#9040, @blackpiglet)
+  * Fix issue #8961, cancel PVB/PVR on Velero server restart (#9031, @Lyndon-Li)
+  * Fix issue #8962, resume PVB/PVR during node-agent restarts (#9030, @Lyndon-Li)
+  * Bump kopia v0.20.1 (#9027, @Lyndon-Li)
+  * Fix issue #8965, support PVB/PVR's cancel state in the backup/restore (#9026, @Lyndon-Li)
+  * Fix Issue 8816 When specifying LabelSelector on restore, related items such as PVC and VolumeSnapshot are not included (#9024, @amastbau)
+  * Fix issue #8963, add legacy PVR controller for Restic path (#9022, @Lyndon-Li)
+  * Fix issue #8964, add Windows support for VGDP MS for fs-backup (#9021, @Lyndon-Li)
+  * Accommodate VGS workflows in PVC CSI plugin (#9019, @shubham-pampattiwar)
+  * Fix issue #8958, add VGDP MS PVB controller (#9015, @Lyndon-Li)
+  * Fix issue #8959, add VGDP MS PVR controller (#9014, @Lyndon-Li)
+  * Fix issue #8988, add data path for VGDP ms PVR (#9005, @Lyndon-Li)
+  * Fix issue #8988, add data path for VGDP ms pvb (#8998, @Lyndon-Li)
+  * Skip VS and VSC not created by backup. (#8990, @blackpiglet)
+  * Make ResticIdentifier optional for kopia BackupRepositories (#8987, @kaovilai)
+  * Fix issue #8960, implement PodVolume exposer for PVB/PVR (#8985, @Lyndon-Li)
+  * fix: update mc command in minio-deployment example (#8982, @vishal-chdhry)
+  * Fix issue #8957, add design for VGDP MS for fs-backup (#8979, @Lyndon-Li)
+  * Add BSL status check for backup/restore operations. (#8976, @blackpiglet)
+  * Mark BackupRepository not ready when BSL changed (#8975, @ywk253100)
+  * Add support for [distributed snapshotting](https://github.com/kubernetes-csi/external-snapshotter/tree/4cedb3f45790ac593ebfa3324c490abedf739477?tab=readme-ov-file#distributed-snapshotting) (#8969, @flx5)
+  * Fix issue #8534, refactor dm controllers to tolerate cancel request in more cases, e.g., node restart, node drain (#8952, @Lyndon-Li)
+  * The backup and restore VGDP affinity enhancement implementation. (#8949, @blackpiglet)
+  * Remove CSI VS and VSC metadata from backup. (#8946, @blackpiglet)
+  * Extend PVCAction itemblock plugin to support grouping PVCs under VGS label key (#8944, @shubham-pampattiwar)
+  * Copy security context from origin pod (#8943, @farodin91)
+  * Add support for configuring VGS label key (#8938, @shubham-pampattiwar)
+  * Add VolumeSnapshotContent into the RIA and the mustHave resource list. (#8924, @blackpiglet)
+  * Mounted cloud credentials should not be world-readable (#8919, @sseago)
+  * Warn for not found error in patching managed fields (#8902, @sseago)
+  * Fix issue 8878, relief node os deduction error checks (#8891, @Lyndon-Li)
+  * Skip namespace in terminating state in backup resource collection. (#8890, @blackpiglet)
+  * Implement PriorityClass Support (#8883, @kaovilai)
+  * Fix Velero adding restore-wait init container when not needed. (#8880, @kaovilai)
+  * Pass the logger in kopia related operations. (#8875, @hu-keyu)
+  * Inherit the dnsPolicy and dnsConfig from the node agent pod. This is done so that the kopia task uses the same configuration. (#8845, @flx5)
+  * Add design for VolumeGroupSnapshot support (#8778, @shubham-pampattiwar)
+  * Inherit k8s default volumeSnapshotClass. (#8719, @hu-keyu)
+  * CLI automatically discovers and uses cacert from BSL for download requests (#8557, @kaovilai)
+  * This PR aims to add s390x support to Velero binary. (#7505, @pandurangkhandeparker)

changelogs/CHANGELOG-1.18.md

@@ -0,0 +1,109 @@
+## v1.18
+
+### Download
+https://github.com/vmware-tanzu/velero/releases/tag/v1.18.0
+
+### Container Image
+`velero/velero:v1.18.0`
+
+### Documentation
+https://velero.io/docs/v1.18/
+
+### Upgrading
+https://velero.io/docs/v1.18/upgrade-to-1.18/
+
+### Highlights
+#### Concurrent backup
+In v1.18, Velero is capable to process multiple backups concurrently. This is a significant usability improvement, especially for multiple tenants or multiple users case, backups submitted from different users could run their backups simultaneously without interfering with each other.  
+
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/Implemented/concurrent-backup-processing.md for more details.
+
+#### Cache volume for data movers
+In v1.18, Velero allows users to configure cache volumes for data mover pods during restore for CSI snapshot data movement and fs-backup. This brings below benefits:
+- Solve the problem that data mover pods fail to when pod's ephemeral disk is limited
+- Solve the problem that multiple data mover pods fail to run concurrently in one node when the node's ephemeral disk is limited
+- Working together with backup repository's cache limit configuration, cache volume with appropriate size helps to improve the restore throughput
+
+Check design https://github.com/vmware-tanzu/velero/blob/main/design/Implemented/backup-repo-cache-volume.md for more details.
+
+#### Incremental size for data movers
+In v1.18, Velero allows users to observe the incremental size of data movers backups for CSI snapshot data movement and fs-backup, so that users could visually see the data reduction due to incremental backup.
+
+#### Wildcard support for namespaces
+In v1.18, Velero allows to use Glob regular expressions for namespace filters during backup and restore, so that users could filter namespaces in a batch manner.
+
+#### VolumePolicy for PVC phase
+In v1.18, Velero VolumePolicy supports actions by PVC phase, which help users to do special operations for PVCs with a specific phase, e.g., skip PVCs in Pending/Lost status from the backup.  
+
+#### Scalability and Resiliency improvements
+##### Prevent Velero server OOM Kill for large backup repositories
+In v1.18, some backup repository operations are delay executed out of Velero server, so Velero server won't be OOM Killed.
+
+#### Performance improvement for VolumePolicy
+In v1.18, VolumePolicy is enhanced for large number of pods/PVCs so that the performance is significantly improved.
+
+#### Events for data mover pod diagnostic
+In v1.18, events are recorded into data mover pod diagnostic, which allows user to see more information for troubleshooting when the data mover pod fails.
+
+### Runtime and dependencies
+Golang runtime: 1.25.7  
+kopia: 0.22.3
+
+### Limitations/Known issues
+
+### Breaking changes
+#### Deprecation of PVC selected node feature
+According to [Velero deprecation policy](https://github.com/vmware-tanzu/velero/blob/main/GOVERNANCE.md#deprecation-policy), PVC selected node feature is deprecated in v1.18. Velero could appropriately handle PVC's selected-node annotation, so users don't need to do anything particularly.
+
+### All Changes
+* Remove backup from running list when backup fails validation (#9498, @sseago)
+* Maintenance Job only uses the first element of the LoadAffinity array (#9494, @blackpiglet)
+* Fix issue #9478, add diagnose info on expose peek fails (#9481, @Lyndon-Li)
+* Add Role, RoleBinding, ClusterRole, and ClusterRoleBinding in restore sequence. (#9474, @blackpiglet)
+* Add maintenance job and data mover pod's labels and annotations setting. (#9452, @blackpiglet)
+* Fix plugin init container names exceeding DNS-1123 limit (#9445, @mpryc)
+* Add PVC-to-Pod cache to improve volume policy performance (#9441, @shubham-pampattiwar)
+* Remove VolumeSnapshotClass from CSI B/R process. (#9431, @blackpiglet)
+* Use hookIndex for recording multiple restore exec hooks. (#9366, @blackpiglet)
+* Sanitize Azure HTTP responses in BSL status messages (#9321, @shubham-pampattiwar)
+* Remove labels associated with previous backups (#9206, @Joeavaikath)
+* Add VolumePolicy support for PVC Phase conditions to allow skipping Pending PVCs (#9166, @claude)
+* feat: Enhance BackupStorageLocation with Secret-based CA certificate support (#9141, @kaovilai)
+* Add `--apply` flag to `install` command, allowing usage of Kubernetes apply to make changes to existing installs (#9132, @mjnagel)
+* Fix issue #9194, add doc for GOMAXPROCS behavior change (#9420, @Lyndon-Li)
+* Apply volume policies to VolumeGroupSnapshot PVC filtering (#9419, @shubham-pampattiwar)
+* Fix issue #9276, add doc for cache volume support (#9418, @Lyndon-Li)
+* Add Prometheus metrics for maintenance jobs (#9414, @shubham-pampattiwar)
+* Fix issue #9400, connect repo first time after creation so that init params could be written (#9407, @Lyndon-Li)
+* Cache volume for PVR (#9397, @Lyndon-Li)
+* Cache volume support for DataDownload (#9391, @Lyndon-Li)
+* don't copy securitycontext from first container if configmap found (#9389, @sseago)
+* Refactor repo provider interface for static configuration (#9379, @Lyndon-Li)
+* Fix issue #9365, prevent fake completion notification due to multiple update of single PVR (#9375, @Lyndon-Li)
+* Add cache volume configuration (#9370, @Lyndon-Li)
+* Track actual resource names for GenerateName in restore status (#9368, @shubham-pampattiwar)
+* Fix managed fields patch for resources using GenerateName (#9367, @shubham-pampattiwar)
+* Support cache volume for generic restore exposer and pod volume exposer (#9362, @Lyndon-Li)
+* Add incrementalSize to DU/PVB for reporting new/changed size (#9357, @sseago)
+* Add snapshotSize for DataDownload, PodVolumeRestore (#9354, @Lyndon-Li)
+* Add cache dir configuration for udmrepo (#9353, @Lyndon-Li)
+* Fix the Job build error when BackupReposiotry name longer than 63. (#9350, @blackpiglet)
+* Add cache configuration to VGDP (#9342, @Lyndon-Li)
+* Fix issue #9332, add bytesDone for cache files (#9333, @Lyndon-Li)
+* Fix typos in documentation (#9329, @T4iFooN-IX)
+* Concurrent backup processing (#9307, @sseago)
+* VerifyJSONConfigs verify every elements in Data. (#9302, @blackpiglet)
+* Fix issue #9267, add events to data mover prepare diagnostic (#9296, @Lyndon-Li)
+* Add option for privileged fs-backup pod (#9295, @sseago)
+* Fix issue #9193, don't connect repo in repo controller (#9291, @Lyndon-Li)
+* Implement concurrency control for cache of native VolumeSnapshotter plugin. (#9281, @0xLeo258)
+* Fix issue #7904, remove the code and doc for PVC node selection (#9269, @Lyndon-Li)
+* Fix schedule controller to prevent backup queue accumulation during extended blocking scenarios by properly handling empty backup phases (#9264, @shubham-pampattiwar)
+* Fix repository maintenance jobs to inherit allowlisted tolerations from Velero deployment (#9256, @shubham-pampattiwar)
+* Implement wildcard namespace pattern expansion for backup namespace includes/excludes. This change adds support for wildcard patterns (*, ?, [abc], {a,b,c}) in namespace includes and excludes during backup operations (#9255, @Joeavaikath)
+* Protect VolumeSnapshot field from race condition during multi-thread backup (#9248, @0xLeo258)
+* Update AzureAD Microsoft Authentication Library to v1.5.0 (#9244, @priyansh17)
+* Get pod list once per namespace in pvc IBA (#9226, @sseago)
+* Fix issue #7725, add design for backup repo cache configuration (#9148, @Lyndon-Li)
+* Fix issue #9229, don't attach backupPVC to the source node (#9233, @Lyndon-Li)
+* feat: Permit specifying annotations for the BackupPVC (#9173, @clementnuss)

changelogs/unreleased/9403-GabriFedi97

@@ -0,0 +1 @@
+Include InitContainer configured as Sidecars when validating the existence of the target containers configured for the Backup Hooks

changelogs/unreleased/9502-Joeavaikath

@@ -0,0 +1 @@
+Support all glob wildcard characters in namespace validation

changelogs/unreleased/9508-kaovilai

@@ -0,0 +1 @@
+Fix VolumePolicy PVC phase condition filter for unbound PVCs (#9507)

changelogs/unreleased/9516-shubham-pampattiwar

@@ -0,0 +1 @@
+Fix VolumeGroupSnapshot restore failure with Ceph RBD CSI driver by creating stub VolumeGroupSnapshotContent during restore and looking up VolumeSnapshotClass by driver for credential support

changelogs/unreleased/9528-Lyndon-Li

@@ -0,0 +1 @@
+Add block data mover design for block level incremental backup by integrating with Kubernetes CBT

changelogs/unreleased/9532-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9343, include PV topology to data mover pod affinities

changelogs/unreleased/9533-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9496, support customized host os

changelogs/unreleased/9540-sseago

@@ -0,0 +1 @@
+Add custom action type to volume policies

changelogs/unreleased/9547-blackpiglet

@@ -0,0 +1 @@
+If BIA return updateObj with SkipFromBackupAnnotation, treat it as skip the resource from backup.

changelogs/unreleased/9554-testsabirweb

@@ -0,0 +1 @@
+Issue #9544: Add test coverage for S3 bucket name in MRAP ARN notation and fix bucket validation to accept ARN format

changelogs/unreleased/9560-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9475, use node-selector instead of nodName for generic restore

changelogs/unreleased/9561-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9460, flush buffer before data mover completes

changelogs/unreleased/9570-H-M-Quang-Ngo

@@ -0,0 +1 @@
+Add schedule_expected_interval_seconds metric for dynamic backup alerting thresholds (#9559)

changelogs/unreleased/9574-blackpiglet

@@ -0,0 +1 @@
+Add ephemeral storage limit and request support for data mover and maintenance job

changelogs/unreleased/9581-shubham-pampattiwar

@@ -0,0 +1 @@
+Fix DBR stuck when CSI snapshot no longer exists in cloud provider

changelogs/unreleased/9614-blackpiglet

@@ -0,0 +1 @@
+Add check for file extraction from tarball.

changelogs/unreleased/9628-priyansh17

@@ -0,0 +1 @@
+Implement original VolumeSnapshotContent deletion for legacy backups

changelogs/unreleased/9634-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9626, let go for uninitialized repo under readonly mode

changelogs/unreleased/9638-adam-jian-zhang

@@ -0,0 +1 @@
+Fix issue #9636, fix configmap lookup in non-default namespaces

changelogs/unreleased/9643-priyansh17

@@ -0,0 +1 @@
+Fix issue #9641, Remove redundant ReadyToUse polling in CSI VolumeSnapshotContent delete plugin

changelogs/unreleased/9653-BassinD

@@ -0,0 +1 @@
+Fix service restore with null healthCheckNodePort in last-applied-configuration label

changelogs/unreleased/9663-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9659, in the case that PVB/PVR/DU/DD is cancelled before the data path is really started, call EndEvent to prevent data mover pod from crashing because of delay event distribution

changelogs/unreleased/9668-adam-jian-zhang

@@ -0,0 +1 @@
+Fix issue #9666, fix node-agent node detection in multiple instances scenario

changelogs/unreleased/9676-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9470, remove restic from repository

changelogs/unreleased/9677-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9469, remove restic for uploader

changelogs/unreleased/9682-adam-jian-zhang

@@ -0,0 +1 @@
+Fix issue #9681, fix restores and podvolumerestores list options to only list in installed namespace

changelogs/unreleased/9683-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9428, increase repo maintenance history queue length from 3 to 25

changelogs/unreleased/9693-priyansh17

@@ -0,0 +1 @@
+Enhance backup deletion logic to handle tarball download failures

changelogs/unreleased/9695-shubham-pampattiwar

@@ -0,0 +1 @@
+Bump external-snapshotter to v8.4.0 and migrate VolumeGroupSnapshot API from v1beta1 to v1beta2 for Kubernetes 1.34+ compatibility

changelogs/unreleased/9700-priyansh17

@@ -0,0 +1 @@
+Fix issue #9699, add a 2-second gap between temporary CSI VolumeSnapshotContent create and delete operations

changelogs/unreleased/9701-emirot

@@ -0,0 +1 @@
+Update Debian base image from bookworm to trixie

changelogs/unreleased/9704-adam-jian-zhang

@@ -0,0 +1 @@
+Fix issue #9703, fix CSI PVC Backup Plugin list options to only list in installed namespace

changelogs/unreleased/9705-emirot

@@ -0,0 +1 @@
+perf: better string concatenation

changelogs/unreleased/9724-Lyndon-Li

@@ -0,0 +1 @@
+Fix issue #9723, extend Unified Repo Interface to support block uploader

changelogs/unreleased/9728-blackpiglet

@@ -0,0 +1 @@
+Remove Restic build from Dockerfile, Makefile and Tiltfile.

config/crd/v1/bases/velero.io_backuprepositories.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: backuprepositories.velero.io
 spec:
   group: velero.io
@@ -54,6 +54,13 @@ spec:
                 description: MaintenanceFrequency is how often maintenance should
                   be run.
                 type: string
+              repositoryConfig:
+                additionalProperties:
+                  type: string
+                description: RepositoryConfig is for repository-specific configuration
+                  fields.
+                nullable: true
+                type: object
               repositoryType:
                 description: RepositoryType indicates the type of the backend repository
                 enum:
@@ -62,9 +69,7 @@ spec:
                 - ""
                 type: string
               resticIdentifier:
-                description: |-
-                  ResticIdentifier is the full restic-compatible string for identifying
-                  this repository.
+                description: Deprecated
                 type: string
               volumeNamespace:
                 description: |-
@@ -74,15 +79,14 @@ spec:
             required:
             - backupStorageLocation
             - maintenanceFrequency
-            - resticIdentifier
             - volumeNamespace
             type: object
           status:
             description: BackupRepositoryStatus is the current status of a BackupRepository.
             properties:
               lastMaintenanceTime:
-                description: LastMaintenanceTime is the last time maintenance was
-                  run.
+                description: LastMaintenanceTime is the last time repo maintenance
+                  succeeded.
                 format: date-time
                 nullable: true
                 type: string
@@ -97,6 +101,33 @@ spec:
                 - Ready
                 - NotReady
                 type: string
+              recentMaintenance:
+                description: RecentMaintenance is status of the recent repo maintenance.
+                items:
+                  properties:
+                    completeTimestamp:
+                      description: CompleteTimestamp is the completion time of the
+                        repo maintenance.
+                      format: date-time
+                      nullable: true
+                      type: string
+                    message:
+                      description: Message is a message about the current status of
+                        the repo maintenance.
+                      type: string
+                    result:
+                      description: Result is the result of the repo maintenance.
+                      enum:
+                      - Succeeded
+                      - Failed
+                      type: string
+                    startTimestamp:
+                      description: StartTimestamp is the start time of the repo maintenance.
+                      format: date-time
+                      nullable: true
+                      type: string
+                  type: object
+                type: array
             type: object
         type: object
     served: true

config/crd/v1/bases/velero.io_backups.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: backups.velero.io
 spec:
   group: velero.io
@@ -63,7 +63,6 @@ spec:
                   DefaultVolumesToRestic specifies whether restic should be used to take a
                   backup of all pod volumes by default.
 
-
                   Deprecated: this field is no longer used and will be removed entirely in future. Use DefaultVolumesToFsBackup instead.
                 nullable: true
                 type: boolean
@@ -176,11 +175,13 @@ spec:
                                     items:
                                       type: string
                                     type: array
+                                    x-kubernetes-list-type: atomic
                                 required:
                                 - key
                                 - operator
                                 type: object
                               type: array
+                              x-kubernetes-list-type: atomic
                             matchLabels:
                               additionalProperties:
                                 type: string
@@ -364,11 +365,13 @@ spec:
                           items:
                             type: string
                           type: array
+                          x-kubernetes-list-type: atomic
                       required:
                       - key
                       - operator
                       type: object
                     type: array
+                    x-kubernetes-list-type: atomic
                   matchLabels:
                     additionalProperties:
                       type: string
@@ -425,11 +428,13 @@ spec:
                             items:
                               type: string
                             type: array
+                            x-kubernetes-list-type: atomic
                         required:
                         - key
                         - operator
                         type: object
                       type: array
+                      x-kubernetes-list-type: atomic
                     matchLabels:
                       additionalProperties:
                         type: string
@@ -502,6 +507,10 @@ spec:
                       uploads to perform when using the uploader.
                     type: integer
                 type: object
+              volumeGroupSnapshotLabelKey:
+                description: VolumeGroupSnapshotLabelKey specifies the label key to
+                  group PVCs under a VGS.
+                type: string
               volumeSnapshotLocations:
                 description: VolumeSnapshotLocations is a list containing names of
                   VolumeSnapshotLocations associated with this backup.
@@ -585,6 +594,8 @@ spec:
                 description: Phase is the current state of the Backup.
                 enum:
                 - New
+                - Queued
+                - ReadyToStart
                 - FailedValidation
                 - InProgress
                 - WaitingForPluginOperations
@@ -616,6 +627,11 @@ spec:
                       filters that happen as items are processed.
                     type: integer
                 type: object
+              queuePosition:
+                description: |-
+                  QueuePosition is the position of the backup in the queue.
+                  Only relevant when Phase is "Queued"
+                type: integer
               startTimestamp:
                 description: |-
                   StartTimestamp records the time a backup was started.

config/crd/v1/bases/velero.io_backupstoragelocations.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: backupstoragelocations.velero.io
 spec:
   group: velero.io
@@ -86,10 +86,13 @@ spec:
                       valid secret key.
                     type: string
                   name:
+                    default: ""
                     description: |-
                       Name of the referent.
+                      This field is effectively required, but due to backwards compatibility is
+                      allowed to be empty. Instances of this type with an empty value here are
+                      almost certainly wrong.
                       More info: https://kubernetes.io/docs/concepts/overview/working-with-objects/names/#names
-                      TODO: Add other useful fields. apiVersion, kind, uid?
                     type: string
                   optional:
                     description: Specify whether the Secret or its key must be defined
@@ -110,10 +113,38 @@ spec:
                     description: Bucket is the bucket to use for object storage.
                     type: string
                   caCert:
-                    description: CACert defines a CA bundle to use when verifying
-                      TLS connections to the provider.
+                    description: |-
+                      CACert defines a CA bundle to use when verifying TLS connections to the provider.
+                      Deprecated: Use CACertRef instead.
                     format: byte
                     type: string
+                  caCertRef:
+                    description: |-
+                      CACertRef is a reference to a Secret containing the CA certificate bundle to use
+                      when verifying TLS connections to the provider. The Secret must be in the same
+                      namespace as the BackupStorageLocation.
+                    properties:
+                      key:
+                        description: The key of the secret to select from.  Must be
+                          a valid secret key.
+                        type: string
+                      name:
+                        default: ""
+                        description: |-
+                          Name of the referent.
+                          This field is effectively required, but due to backwards compatibility is
+                          allowed to be empty. Instances of this type with an empty value here are
+                          almost certainly wrong.
+                          More info: https://kubernetes.io/docs/concepts/overview/working-with-objects/names/#names
+                        type: string
+                      optional:
+                        description: Specify whether the Secret or its key must be
+                          defined
+                        type: boolean
+                    required:
+                    - key
+                    type: object
+                    x-kubernetes-map-type: atomic
                   prefix:
                     description: Prefix is the path inside a bucket to use for Velero
                       storage. Optional.
@@ -141,7 +172,6 @@ spec:
                 description: |-
                   AccessMode is an unused field.
 
-
                   Deprecated: there is now an AccessMode field on the Spec and this field
                   will be removed entirely as of v2.0.
                 enum:
@@ -153,7 +183,6 @@ spec:
                   LastSyncedRevision is the value of the `metadata/revision` file in the backup
                   storage location the last time the BSL's contents were synced into the cluster.
 
-
                   Deprecated: this field is no longer updated or used for detecting changes to
                   the location's contents and will be removed entirely in v2.0.
                 type: string

config/crd/v1/bases/velero.io_deletebackuprequests.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: deletebackuprequests.velero.io
 spec:
   group: velero.io

config/crd/v1/bases/velero.io_downloadrequests.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: downloadrequests.velero.io
 spec:
   group: velero.io

config/crd/v1/bases/velero.io_podvolumebackups.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: podvolumebackups.velero.io
 spec:
   group: velero.io
@@ -15,38 +15,47 @@ spec:
   scope: Namespaced
   versions:
   - additionalPrinterColumns:
-    - description: Pod Volume Backup status such as New/InProgress
+    - description: PodVolumeBackup status such as New/InProgress
       jsonPath: .status.phase
       name: Status
       type: string
-    - description: Time when this backup was started
+    - description: Time duration since this PodVolumeBackup was started
       jsonPath: .status.startTimestamp
-      name: Created
+      name: Started
       type: date
-    - description: Namespace of the pod containing the volume to be backed up
-      jsonPath: .spec.pod.namespace
-      name: Namespace
-      type: string
-    - description: Name of the pod containing the volume to be backed up
-      jsonPath: .spec.pod.name
-      name: Pod
-      type: string
-    - description: Name of the volume to be backed up
-      jsonPath: .spec.volume
-      name: Volume
-      type: string
-    - description: The type of the uploader to handle data transfer
-      jsonPath: .spec.uploaderType
-      name: Uploader Type
-      type: string
+    - description: Completed bytes
+      format: int64
+      jsonPath: .status.progress.bytesDone
+      name: Bytes Done
+      type: integer
+    - description: Total bytes
+      format: int64
+      jsonPath: .status.progress.totalBytes
+      name: Total Bytes
+      type: integer
+    - description: Incremental bytes
+      format: int64
+      jsonPath: .status.incrementalBytes
+      name: Incremental Bytes
+      priority: 10
+      type: integer
     - description: Name of the Backup Storage Location where this backup should be
         stored
       jsonPath: .spec.backupStorageLocation
       name: Storage Location
       type: string
-    - jsonPath: .metadata.creationTimestamp
+    - description: Time duration since this PodVolumeBackup was created
+      jsonPath: .metadata.creationTimestamp
       name: Age
       type: date
+    - description: Name of the node where the PodVolumeBackup is processed
+      jsonPath: .status.node
+      name: Node
+      type: string
+    - description: The type of the uploader to handle data transfer
+      jsonPath: .spec.uploaderType
+      name: Uploader
+      type: string
     name: v1
     schema:
       openAPIV3Schema:
@@ -76,6 +85,11 @@ spec:
                   BackupStorageLocation is the name of the backup storage location
                   where the backup repository is stored.
                 type: string
+              cancel:
+                description: |-
+                  Cancel indicates request to cancel the ongoing PodVolumeBackup. It can be set
+                  when the PodVolumeBackup is in InProgress phase
+                type: boolean
               node:
                 description: Node is the name of the node that the Pod is running
                   on.
@@ -96,7 +110,6 @@ spec:
                       the event) or if no container name is specified "spec.containers[2]" (container with
                       index 2 in this pod). This syntax is chosen only to have some well-defined way of
                       referencing a part of an object.
-                      TODO: this design is not final and this field is subject to change in the future.
                     type: string
                   kind:
                     description: |-
@@ -166,6 +179,13 @@ spec:
           status:
             description: PodVolumeBackupStatus is the current status of a PodVolumeBackup.
             properties:
+              acceptedTimestamp:
+                description: |-
+                  AcceptedTimestamp records the time the pod volume backup is to be prepared.
+                  The server's time is used for AcceptedTimestamp
+                format: date-time
+                nullable: true
+                type: string
               completionTimestamp:
                 description: |-
                   CompletionTimestamp records the time a backup was completed.
@@ -175,6 +195,11 @@ spec:
                 format: date-time
                 nullable: true
                 type: string
+              incrementalBytes:
+                description: IncrementalBytes holds the number of bytes new or changed
+                  since the last backup
+                format: int64
+                type: integer
               message:
                 description: Message is a message about the pod volume backup's status.
                 type: string
@@ -186,7 +211,11 @@ spec:
                 description: Phase is the current state of the PodVolumeBackup.
                 enum:
                 - New
+                - Accepted
+                - Prepared
                 - InProgress
+                - Canceling
+                - Canceled
                 - Completed
                 - Failed
                 type: string

config/crd/v1/bases/velero.io_podvolumerestores.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: podvolumerestores.velero.io
 spec:
   group: velero.io
@@ -15,39 +15,40 @@ spec:
   scope: Namespaced
   versions:
   - additionalPrinterColumns:
-    - description: Namespace of the pod containing the volume to be restored
-      jsonPath: .spec.pod.namespace
-      name: Namespace
+    - description: PodVolumeRestore status such as New/InProgress
+      jsonPath: .status.phase
+      name: Status
       type: string
-    - description: Name of the pod containing the volume to be restored
-      jsonPath: .spec.pod.name
-      name: Pod
+    - description: Time duration since this PodVolumeRestore was started
+      jsonPath: .status.startTimestamp
+      name: Started
+      type: date
+    - description: Completed bytes
+      format: int64
+      jsonPath: .status.progress.bytesDone
+      name: Bytes Done
+      type: integer
+    - description: Total bytes
+      format: int64
+      jsonPath: .status.progress.totalBytes
+      name: Total Bytes
+      type: integer
+    - description: Name of the Backup Storage Location where the backup data is stored
+      jsonPath: .spec.backupStorageLocation
+      name: Storage Location
+      type: string
+    - description: Time duration since this PodVolumeRestore was created
+      jsonPath: .metadata.creationTimestamp
+      name: Age
+      type: date
+    - description: Name of the node where the PodVolumeRestore is processed
+      jsonPath: .status.node
+      name: Node
       type: string
     - description: The type of the uploader to handle data transfer
       jsonPath: .spec.uploaderType
       name: Uploader Type
       type: string
-    - description: Name of the volume to be restored
-      jsonPath: .spec.volume
-      name: Volume
-      type: string
-    - description: Pod Volume Restore status such as New/InProgress
-      jsonPath: .status.phase
-      name: Status
-      type: string
-    - description: Pod Volume Restore status such as New/InProgress
-      format: int64
-      jsonPath: .status.progress.totalBytes
-      name: TotalBytes
-      type: integer
-    - description: Pod Volume Restore status such as New/InProgress
-      format: int64
-      jsonPath: .status.progress.bytesDone
-      name: BytesDone
-      type: integer
-    - jsonPath: .metadata.creationTimestamp
-      name: Age
-      type: date
     name: v1
     schema:
       openAPIV3Schema:
@@ -77,6 +78,11 @@ spec:
                   BackupStorageLocation is the name of the backup storage location
                   where the backup repository is stored.
                 type: string
+              cancel:
+                description: |-
+                  Cancel indicates request to cancel the ongoing PodVolumeRestore. It can be set
+                  when the PodVolumeRestore is in InProgress phase
+                type: boolean
               pod:
                 description: Pod is a reference to the pod containing the volume to
                   be restored.
@@ -93,7 +99,6 @@ spec:
                       the event) or if no container name is specified "spec.containers[2]" (container with
                       index 2 in this pod). This syntax is chosen only to have some well-defined way of
                       referencing a part of an object.
-                      TODO: this design is not final and this field is subject to change in the future.
                     type: string
                   kind:
                     description: |-
@@ -128,6 +133,10 @@ spec:
               snapshotID:
                 description: SnapshotID is the ID of the volume snapshot to be restored.
                 type: string
+              snapshotSize:
+                description: SnapshotSize is the logical size in Bytes of the snapshot.
+                format: int64
+                type: integer
               sourceNamespace:
                 description: SourceNamespace is the original namespace for namaspace
                   mapping.
@@ -163,6 +172,13 @@ spec:
           status:
             description: PodVolumeRestoreStatus is the current status of a PodVolumeRestore.
             properties:
+              acceptedTimestamp:
+                description: |-
+                  AcceptedTimestamp records the time the pod volume restore is to be prepared.
+                  The server's time is used for AcceptedTimestamp
+                format: date-time
+                nullable: true
+                type: string
               completionTimestamp:
                 description: |-
                   CompletionTimestamp records the time a restore was completed.
@@ -174,11 +190,19 @@ spec:
               message:
                 description: Message is a message about the pod volume restore's status.
                 type: string
+              node:
+                description: Node is name of the node where the pod volume restore
+                  is processed.
+                type: string
               phase:
                 description: Phase is the current state of the PodVolumeRestore.
                 enum:
                 - New
+                - Accepted
+                - Prepared
                 - InProgress
+                - Canceling
+                - Canceled
                 - Completed
                 - Failed
                 type: string

config/crd/v1/bases/velero.io_restores.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: restores.velero.io
 spec:
   group: velero.io
@@ -138,11 +138,13 @@ spec:
                                     items:
                                       type: string
                                     type: array
+                                    x-kubernetes-list-type: atomic
                                 required:
                                 - key
                                 - operator
                                 type: object
                               type: array
+                              x-kubernetes-list-type: atomic
                             matchLabels:
                               additionalProperties:
                                 type: string
@@ -291,11 +293,13 @@ spec:
                           items:
                             type: string
                           type: array
+                          x-kubernetes-list-type: atomic
                       required:
                       - key
                       - operator
                       type: object
                     type: array
+                    x-kubernetes-list-type: atomic
                   matchLabels:
                     additionalProperties:
                       type: string
@@ -354,11 +358,13 @@ spec:
                             items:
                               type: string
                             type: array
+                            x-kubernetes-list-type: atomic
                         required:
                         - key
                         - operator
                         type: object
                       type: array
+                      x-kubernetes-list-type: atomic
                     matchLabels:
                       additionalProperties:
                         type: string

config/crd/v1/bases/velero.io_schedules.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: schedules.velero.io
 spec:
   group: velero.io
@@ -102,7 +102,6 @@ spec:
                       DefaultVolumesToRestic specifies whether restic should be used to take a
                       backup of all pod volumes by default.
 
-
                       Deprecated: this field is no longer used and will be removed entirely in future. Use DefaultVolumesToFsBackup instead.
                     nullable: true
                     type: boolean
@@ -215,11 +214,13 @@ spec:
                                         items:
                                           type: string
                                         type: array
+                                        x-kubernetes-list-type: atomic
                                     required:
                                     - key
                                     - operator
                                     type: object
                                   type: array
+                                  x-kubernetes-list-type: atomic
                                 matchLabels:
                                   additionalProperties:
                                     type: string
@@ -405,11 +406,13 @@ spec:
                               items:
                                 type: string
                               type: array
+                              x-kubernetes-list-type: atomic
                           required:
                           - key
                           - operator
                           type: object
                         type: array
+                        x-kubernetes-list-type: atomic
                       matchLabels:
                         additionalProperties:
                           type: string
@@ -466,11 +469,13 @@ spec:
                                 items:
                                   type: string
                                 type: array
+                                x-kubernetes-list-type: atomic
                             required:
                             - key
                             - operator
                             type: object
                           type: array
+                          x-kubernetes-list-type: atomic
                         matchLabels:
                           additionalProperties:
                             type: string
@@ -544,6 +549,10 @@ spec:
                           uploads to perform when using the uploader.
                         type: integer
                     type: object
+                  volumeGroupSnapshotLabelKey:
+                    description: VolumeGroupSnapshotLabelKey specifies the label key
+                      to group PVCs under a VGS.
+                    type: string
                   volumeSnapshotLocations:
                     description: VolumeSnapshotLocations is a list containing names
                       of VolumeSnapshotLocations associated with this backup.

config/crd/v1/bases/velero.io_serverstatusrequests.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: serverstatusrequests.velero.io
 spec:
   group: velero.io

config/crd/v1/bases/velero.io_volumesnapshotlocations.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: volumesnapshotlocations.velero.io
 spec:
   group: velero.io
@@ -57,10 +57,13 @@ spec:
                       valid secret key.
                     type: string
                   name:
+                    default: ""
                     description: |-
                       Name of the referent.
+                      This field is effectively required, but due to backwards compatibility is
+                      allowed to be empty. Instances of this type with an empty value here are
+                      almost certainly wrong.
                       More info: https://kubernetes.io/docs/concepts/overview/working-with-objects/names/#names
-                      TODO: Add other useful fields. apiVersion, kind, uid?
                     type: string
                   optional:
                     description: Specify whether the Secret or its key must be defined

config/crd/v2alpha1/bases/velero.io_datadownloads.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: datadownloads.velero.io
 spec:
   group: velero.io
@@ -92,6 +92,13 @@ spec:
                   DataMover specifies the data mover to be used by the backup.
                   If DataMover is "" or "velero", the built-in data mover will be used.
                 type: string
+              nodeOS:
+                description: NodeOS is OS of the node where the DataDownload is processed.
+                enum:
+                - auto
+                - linux
+                - windows
+                type: string
               operationTimeout:
                 description: |-
                   OperationTimeout specifies the time used to wait internal operations,
@@ -101,6 +108,10 @@ spec:
                 description: SnapshotID is the ID of the Velero backup snapshot to
                   be restored from.
                 type: string
+              snapshotSize:
+                description: SnapshotSize is the logical size in Bytes of the snapshot.
+                format: int64
+                type: integer
               sourceNamespace:
                 description: |-
                   SourceNamespace is the original namespace where the volume is backed up from.
@@ -136,6 +147,16 @@ spec:
           status:
             description: DataDownloadStatus is the current status of a DataDownload.
             properties:
+              acceptedByNode:
+                description: Node is name of the node where the DataUpload is prepared.
+                type: string
+              acceptedTimestamp:
+                description: |-
+                  AcceptedTimestamp records the time the DataUpload is to be prepared.
+                  The server's time is used for AcceptedTimestamp
+                format: date-time
+                nullable: true
+                type: string
               completionTimestamp:
                 description: |-
                   CompletionTimestamp records the time a restore was completed.

config/crd/v2alpha1/bases/velero.io_datauploads.yaml

@@ -3,7 +3,7 @@ apiVersion: apiextensions.k8s.io/v1
 kind: CustomResourceDefinition
 metadata:
   annotations:
-    controller-gen.kubebuilder.io/version: v0.14.0
+    controller-gen.kubebuilder.io/version: v0.16.5
   name: datauploads.velero.io
 spec:
   group: velero.io
@@ -33,6 +33,12 @@ spec:
       jsonPath: .status.progress.totalBytes
       name: Total Bytes
       type: integer
+    - description: Incremental bytes
+      format: int64
+      jsonPath: .status.incrementalBytes
+      name: Incremental Bytes
+      priority: 10
+      type: integer
     - description: Name of the Backup Storage Location where this backup should be
         stored
       jsonPath: .spec.backupStorageLocation
@@ -87,6 +93,9 @@ spec:
                   of the CSI snapshot.
                 nullable: true
                 properties:
+                  driver:
+                    description: Driver is the driver used by the VolumeSnapshotContent
+                    type: string
                   snapshotClass:
                     description: SnapshotClass is the name of the snapshot class that
                       the volume snapshot is created with
@@ -143,6 +152,17 @@ spec:
           status:
             description: DataUploadStatus is the current status of a DataUpload.
             properties:
+              acceptedByNode:
+                description: AcceptedByNode is name of the node where the DataUpload
+                  is prepared.
+                type: string
+              acceptedTimestamp:
+                description: |-
+                  AcceptedTimestamp records the time the DataUpload is to be prepared.
+                  The server's time is used for AcceptedTimestamp
+                format: date-time
+                nullable: true
+                type: string
               completionTimestamp:
                 description: |-
                   CompletionTimestamp records the time a backup was completed.
@@ -159,12 +179,24 @@ spec:
                   as a result of the DataUpload.
                 nullable: true
                 type: object
+              incrementalBytes:
+                description: IncrementalBytes holds the number of bytes new or changed
+                  since the last backup
+                format: int64
+                type: integer
               message:
                 description: Message is a message about the DataUpload's status.
                 type: string
               node:
                 description: Node is name of the node where the DataUpload is processed.
                 type: string
+              nodeOS:
+                description: NodeOS is OS of the node where the DataUpload is processed.
+                enum:
+                - auto
+                - linux
+                - windows
+                type: string
               path:
                 description: Path is the full path of the snapshot volume being backed
                   up.

config/rbac/role.yaml

@@ -8,17 +8,7 @@ rules:
   - ""
   resources:
   - persistentvolumerclaims
-  verbs:
-  - get
-- apiGroups:
-  - ""
-  resources:
   - persistentvolumes
-  verbs:
-  - get
-- apiGroups:
-  - ""
-  resources:
   - pods
   verbs:
   - get
@@ -26,6 +16,18 @@ rules:
   - velero.io
   resources:
   - backuprepositories
+  - backups
+  - backupstoragelocations
+  - datadownloads
+  - datauploads
+  - deletebackuprequests
+  - downloadrequests
+  - podvolumebackups
+  - podvolumerestores
+  - restores
+  - schedules
+  - serverstatusrequests
+  - volumesnapshotlocations
   verbs:
   - create
   - delete
@@ -38,239 +40,18 @@ rules:
   - velero.io
   resources:
   - backuprepositories/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - backups
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - backups/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - backupstoragelocations
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - backupstoragelocations/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - datadownloads
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - datadownloads/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - datauploads
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - datauploads/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - deletebackuprequests
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - deletebackuprequests/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - downloadrequests
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - downloadrequests/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - podvolumebackups
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - podvolumebackups/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - podvolumerestores
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - podvolumerestores/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - restores
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - restores/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - schedules
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - schedules/status
-  verbs:
-  - get
-  - patch
-  - update
-- apiGroups:
-  - velero.io
-  resources:
-  - serverstatusrequests
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch
-- apiGroups:
-  - velero.io
-  resources:
   - serverstatusrequests/status
   verbs:
   - get
   - patch
   - update
-- apiGroups:
-  - velero.io
-  resources:
-  - volumesnapshotlocations
-  verbs:
-  - create
-  - delete
-  - get
-  - list
-  - patch
-  - update
-  - watch

design/Implemented/Extend-VolumePolicies-to-support-more-actions.md

@@ -76,7 +76,7 @@ volumePolicies:
 - Update VolumePolicy action type validation to account for `fs-backup` and `snapshot` as valid VolumePolicy actions.
 - Modifications needed for `fs-backup` action:
   - Now based on the specification of volume policy on backup request we will decide whether to go via legacy pod annotations approach or the newer volume policy based fs-backup action approach.
-  - If there is a presence of volume policy(fs-backup/snapshot)  on the backup request that matches as an action for a volume we use the newer volume policy approach to get the list of the volumes for `fs-backup` action
+  - If there is a presence of volume policy(fs-backup/snapshot) on the backup request that matches as an action for a volume we use the newer volume policy approach to get the list of the volumes for `fs-backup` action
   - Else continue with the annotation based legacy approach workflow.
 
 - Modifications needed for `snapshot` action:
@@ -276,7 +276,7 @@ func (v *volumeHelperImpl) ShouldPerformSnapshot(obj runtime.Unstructured, group
 
 	if !boolptr.IsSetToFalse(v.snapshotVolumes) {
 		// If the backup.Spec.SnapshotVolumes is not set, or set to true, then should take the snapshot.
-		v.logger.Infof("performing snapshot action for pv %s as the snapshotVolumes is not set to false")
+		v.logger.Infof("performing snapshot action for pv %s as the snapshotVolumes is not set to false", pv.Name)
 		return true, nil
 	}
 

design/Implemented/apply-flag.md

@@ -0,0 +1,70 @@
+# Apply flag for install command
+
+## Abstract
+Add an `--apply` flag to the install command that enables applying existing resources rather than creating them. This can be useful as part of the upgrade process for existing installations.
+
+## Background
+The current Velero install command creates resources but doesn't provide a direct way to apply updates to an existing installation.
+Users attempting to run the install command on an existing installation receive "already exists" messages.
+Upgrade steps for existing installs typically involve a three (or more) step process to apply updated CRDs (using `--dry-run` and piping to `kubectl apply`) and then updating/setting images on the Velero deployment and node-agent.
+
+## Goals
+- Provide a simple flag to enable applying resources on an existing Velero installation.
+- Use server-side apply to update existing resources rather than attempting to create them.
+- Maintain consistency with the regular install flow.
+
+## Non Goals
+- Implement special logic for specific version-to-version upgrades (i.e. resource deletion, etc).
+- Add complex upgrade validation or pre/post-upgrade hooks.
+- Provide rollback capabilities.
+
+## High-Level Design
+The `--apply` flag will be added to the Velero install command.
+When this flag is set, the installation process will use server-side apply to update existing resources instead of using create on new resources.
+This flag can be used as _part_ of the upgrade process, but will not always fully handle an upgrade.
+
+## Detailed Design
+The implementation adds a new boolean flag `--apply` to the install command.
+This flag will be passed through to the underlying install functions where the resource creation logic resides.
+
+When the flag is set to true:
+- The `createOrApplyResource` function will use server-side apply with field manager "velero-cli" and `force=true` to update resources.
+- Resources will be applied in the same order as they would be created during installation.
+- Custom Resource Definitions will still be processed first, and the system will wait for them to be established before continuing.
+
+The server-side apply approach with `force=true` ensures that resources are updated even if there are conflicts with the last applied state.
+This provides a best-effort mechanism to apply resources that follows the same flow as installation but updates resources instead of creating them.
+
+No special handling is added for specific versions or resource structures, making this a general-purpose mechanism for applying resources.
+
+## Alternatives Considered
+1. Creating a separate `upgrade` command that would duplicate much of the install command logic.
+   - Rejected due to code duplication and maintenance overhead.
+
+2. Implementing version-specific upgrade logic to handle breaking changes between versions.
+   - Rejected as overly complex and difficult to maintain across multiple version paths.
+   - This could be considered again in the future, but is not in the scope of the current design.
+
+3. Adding automatic detection of existing resources and switching to apply mode.
+   - Rejected as it could lead to unexpected behavior and confusion if users unintentionally apply changes to existing resources.
+
+## Security Considerations
+The apply flag maintains the same security profile as the install command.
+No additional permissions are required beyond what is needed for resource creation.
+The use of `force=true` with server-side apply could potentially override manual changes made to resources, but this is a necessary trade-off to ensure apply is successful.
+
+## Compatibility
+This enhancement is compatible with all existing Velero installations as it is a new opt-in flag.
+It does not change any resource formats or API contracts.
+The apply process is best-effort and does not guarantee compatibility between arbitrary versions of Velero.
+Users should still consult release notes for any breaking changes that may require manual intervention.
+This flag could be adopted by the helm chart, specifically for CRD updates, to simplify the CRD update job.
+
+## Implementation
+The implementation involves:
+1. Adding support for `Apply` to the existing Kubernetes client code.
+1. Adding the `--apply` flag to the install command options.
+1. Changing `createResource` to `createOrApplyResource` and updating it to use server-side apply when the `apply` boolean is set.
+
+The implementation is straightforward and follows existing code patterns.
+No migration of state or special handling of specific resources is required.

design/Implemented/backup-performance-improvements.md

@@ -0,0 +1,370 @@
+# Velero Backup performance Improvements and VolumeGroupSnapshot enablement
+
+There are two different goals here, linked by a single primary missing feature in the Velero backup workflow.
+The first goal is to enhance backup performance by allowing the primary backup controller to run in multiple threads, enabling Velero to back up multiple items at the same time for a given backup.
+The second goal is to enable Velero to eventually support VolumeGroupSnapshots.
+For both of these goals, Velero needs a way to determine which items should be backed up together.
+
+This design proposal will include two development phases:
+- Phase 1 will refactor the backup workflow to identify blocks of related items that should be backed up together, and then coordinate backup hooks among items in the block.
+- Phase 2 will add multiple worker threads for backing up item blocks, so instead of backing up each block as it identified, the velero backup workflow will instead add the block to a channel and one of the workers will pick it up.
+- Actual support for VolumeGroupSnapshots is out-of-scope here and will be handled in a future design proposal, but the item block refactor introduced in Phase 1 is a primary building block for this future proposal.
+
+## Background
+Currently, during backup processing, the main Velero backup controller runs in a single thread, completely finishing the primary backup processing for one resource before moving on to the next one.
+We can improve the overall backup performance by backing up multiple items for a backup at the same time, but before we can do this we must first identify resources that need to be backed up together.
+Generally speaking, resources that need to be backed up together are resources with interdependencies -- pods with their PVCs, PVCs with their PVs, groups of pods that form a single application, CRs, pods, and other resources that belong to the same operator, etc.
+As part of this initial refactoring, once these "Item Blocks" are identified, an additional change will be to move pod hook processing up to the ItemBlock level.
+If there are multiple pods in the ItemBlock, pre-hooks for all pods will be run before backing up the items, followed by post-hooks for all pods.
+This change to hook processing is another prerequisite for future VolumeGroupSnapshot support, since supporting this will require backing up the pods and volumes together for any volumes which belong to the same group.
+Once we are backing up items by block, the next step will be to create multiple worker threads to process and back up ItemBlocks, so that we can back up multiple ItemBlocks at the same time.
+
+In looking at the different kinds of large backups that Velero must deal with, two obvious scenarios come to mind:
+1. Backups with a relatively small number of large volumes
+2. Backups with a large number of relatively small volumes.
+
+In case 1, the majority of the time spent on the backup is in the asynchronous phases -- CSI snapshot creation actions after the snaphandle exists, and DataUpload processing. In that case, parallel item processing will likely have a minimal impact on overall backup completion time.
+
+In case 2, the majority of time spent on the backup will likely be during the synchronous actions. Especially as regards CSI snapshot creation, the waiting for the VSC snaphandle to exist will result in significant passage of time with thousands of volumes. This is the sort of use case which will benefit the most from parallel item processing.
+
+## Goals
+- Identify groups of related items to back up together (ItemBlocks).
+- Manage backup hooks at the ItemBlock level rather than per-item.
+- Using worker threads, back up ItemBlocks at the same time.
+
+## Non Goals
+- Support VolumeGroupSnapshots: this is a future feature, although certain prerequisites for this enhancement are included in this proposal.
+- Process multiple backups in parallel: this is a future feature, although certain prerequisites for this enhancement are included in this proposal.
+- Refactoring plugin infrastructure to avoid RPC calls for internal plugins.
+- Restore performance improvements: this is potentially a future feature
+
+## High-Level Design
+
+### ItemBlock concept
+
+The updated design is based on a new struct/type called `ItemBlock`.
+Essentially, an `ItemBlock` is a group of items that must be backed up together in order to guarantee backup integrity.
+When we eventually split item backup across multiple worker threads, `ItemBlocks` will be kept together as the basic unit of backup.
+To facilitate this, a new plugin type, `ItemBlockAction` will allow relationships between items to be identified by velero -- any resources that must be backed up with other resources will need IBA plugins defined for them.
+Examples of `ItemBlocks` include:
+1. A pod, its mounted PVCs, and the bound PVs for those PVCs.
+2. A VolumeGroup (related PVCs and PVs) along with any pods mounting these volumes.
+3. For a ReadWriteMany PVC, the PVC, its bound PV, and all pods mounting this PVC.
+
+### Phase 1: ItemBlock processing
+- A new plugin type, `ItemBlockAction`, will be created
+- `ItemBlockAction` will contain the API method `GetRelatedItems`, which will be needed for determining which items to group together into `ItemBlocks`.
+- When processing the list of items returned from the item collector, instead of simply calling `BackupItem` on each in turn, we will use the `GetRelatedItems` API call to determine other items to include with the current item in an ItemBlock. Repeat recursively on each item returned.
+- Don't include an item in more than one ItemBlock -- if the next item from the item collector is already in a block, skip it.
+- Once ItemBlock is determined, call new func `BackupItemBlock` instead of `BackupItem`.
+- New func `BackupItemBlock` will call pre hooks for any pods in the block, then back up the items in the block (`BackupItem` will no longer run hooks directly), then call post hooks for any pods in the block.
+- The finalize phase will not be affected by the ItemBlock design, since this is just updating resources after async operations are completed on the items and there is no need to run these updates in parallel.
+
+### Phase 2: Process ItemBlocks for a single backup in multiple threads
+- Concurrent `BackupItemBlock` operations will be executed by worker threads invoked by the backup controller, which will communicate with the backup controller operation via a shared channel.
+- The ItemBlock processing loop implemented in Phase 1 will be modified to send each newly-created ItemBlock to the shared channel rather than calling `BackupItemBlock` inline.
+- Users will be able to configure the number of workers available for concurrent `BackupItemBlock` operations.
+- Access to the BackedUpItems map must be synchronized
+
+## Detailed Design
+
+### Phase 1: ItemBlock processing
+
+#### New ItemBlockAction plugin type
+
+In order for Velero to identify groups of items to back up together in an ItemBlock, we need a way to identify items which need to be backed up along with the current item. While the current `Execute` BackupItemAction method does return a list of additional items which are required by the current item, we need to know this *before* we start the item backup. To support this, we need a new plugin type, `ItemBlockAction` (IBA) with an API method, `GetRelatedItems` which Velero will call on each item as it processes. The expectation is that the registered IBA plugins will return the same items as returned as additional items by the BIA `Execute` method, with the exception that items which are not created until calling `Execute` should not be returned here, as they don't exist yet.
+
+#### Proto definition (compiled into golang by protoc)
+
+The ItemBlockAction plugin type is defined as follows:
+```
+service ItemBlockAction {
+    rpc AppliesTo(ItemBlockActionAppliesToRequest) returns (ItemBlockActionAppliesToResponse);
+    rpc GetRelatedItems(ItemBlockActionGetRelatedItemsRequest) returns (ItemBlockActionGetRelatedItemsResponse);
+}
+
+message ItemBlockActionAppliesToRequest {
+    string plugin = 1;
+}
+
+message ItemBlockActionAppliesToResponse {
+    ResourceSelector ResourceSelector = 1;
+}
+
+message ItemBlockActionGetRelatedItemsRequest {
+    string plugin = 1;
+    bytes item = 2;
+    bytes backup = 3;
+}
+
+message ItemBlockActionGetRelatedItemsResponse {
+    repeated generated.ResourceIdentifier relatedItems = 1;
+}
+```
+
+A new PluginKind, `ItemBlockAction`, will be created, and the backup process will be modified to use this plugin kind.
+
+For any BIA plugins which return additional items from `Execute()` that need to be backed up at the same time or sequentially in the same worker thread as the current items should add a new IBA plugin to return these same items (minus any which won't exist before BIA `Execute()` is called).
+This mainly applies to plugins that operate on pods which reference resources which must be backed up along with the pod and are potentially affected by pod hooks or for plugins which connect multiple pods whose volumes should be backed up at the same time.
+
+### Changes to processing item list from the Item Collector
+
+#### New structs BackupItemBlock, ItemBlock, and ItemBlockItem
+```go
+package backup
+
+type BackupItemBlock struct {
+    itemblock.ItemBlock
+    // This is a reference to the  shared itemBackupper for the backup
+    itemBackupper *itemBackupper
+}
+
+package itemblock
+
+type ItemBlock struct {
+    Log           logrus.FieldLogger
+    Items         []ItemBlockItem
+}
+
+type ItemBlockItem struct {
+    Gr           schema.GroupResource
+    Item         *unstructured.Unstructured
+    PreferredGVR schema.GroupVersionResource
+}
+```
+
+#### Current workflow
+In the `BackupWithResolvers` func, the current Velero implementation iterates over the list of items for backup returned by the Item Collector. For each item, Velero loads the item from the file created by the Item Collector, we call `backupItem`, update the GR map if successful, remove the (temporary) file containing item metadata, and update progress for the backup.
+
+#### Modifications to the loop over ItemCollector results
+The `kubernetesResource` struct used by the item collector will be modified to add an `orderedResource` bool which will be set true for all of the resources moved to the beginning for each GroupResource as a result of being ordered resources.
+In addition, an `inItemBlock` bool is added to the struct which will be set to true later when processing the list when each item is added to an ItemBlock.
+While the item collector already puts ordered resources first for each GR, there is no indication in the list which of these initial items are from the ordered resources list and which are the remaining (unordered) items.
+Velero needs to know which resources are ordered because when we process them later, the ordered resources for each GroupResource must be processed sequentially in a single ItemBlock.
+
+The current workflow within each iteration of the ItemCollector.items loop will replaced with the following:
+- (note that some of the below should be pulled out into a helper func to facilitate recursive call to it for items returned from `GetRelatedItems`.)
+- Before loop iteration, create a pointer to a `BackupItemBlock` which will represent the current ItemBlock being processed.
+- If `item` has `inItemBlock==true`, continue. This one has already been processed.
+- If current `itemBlock` is nil, create it.
+- Add `item` to `itemBlock`.
+- Load item from ItemCollector file. Close/remove file after loading (on error return or not, possibly with similar anonymous func to current impl)
+- If other versions of the same item exist (via EnableAPIGroupVersions), add these to the `itemBlock` as well (and load from ItemCollector file)
+- Get matching IBA plugins for item, call `GetRelatedItems` for each. For each item returned, get full item content from ItemCollector (if present in item list, pulling from file, removing file when done) or from cluster (if not present in item list), add item to the current block, add item to `itemsInBlock` map, and then recursively apply current step to each (i.e. call IBA method, add to block, etc.)
+- If current item and next item are both ordered items for the same GR, then continue to next item, adding to current `itemBlock`.
+- Once full ItemBlock list is generated, call `backupItemBlock(block ItemBlock)
+- Add `backupItemBlock` return values to `backedUpGroupResources` map
+
+
+#### New func `backupItemBlock`
+
+Method signature for new func `backupItemBlock` is as follows:
+```go
+func (kb *kubernetesBackupper) backupItemBlock(block BackupItemBlock) []schema.GroupResource
+```
+The return value is a slice of GRs for resources which were backed up. Velero tracks these to determine which CRDs need to be included in the backup. Note that we need to make sure we include in this not only those resources that were backed up directly, but also those backed up indirectly via additional items BIA execute returns.
+
+In order to handle backup hooks, this func will first take the input item list (`block.items`) and get a list of included pods, filtered to include only those not yet backed up (using `block.itemBackupper.backupRequest.BackedUpItems`). Iterate over this list and execute pre hooks (pulled out of `itemBackupper.backupItemInternal`) for each item.
+Now iterate over the full list (`block.items`) and call `backupItem` for each. After the first, the later items should already have been backed up, but calling a second time is harmless, since the first thing Velero does is check the `BackedUpItems` map, exiting if item is already backed up). We still need this call in case there's a plugin which returns something in `GetAdditionalItems` but forgets to return it in the `Execute` additional items return value. If we don't do this, we could end up missing items.
+
+After backing up the items in the block, we now execute post hooks using the same filtered item list we used for pre hooks, again taking the logic from `itemBackupper.backupItemInternal`).
+
+#### `itemBackupper.backupItemInternal` cleanup
+
+After implementing backup hooks in `backupItemBlock`, hook processing should be removed from `itemBackupper.backupItemInternal`.
+
+### Phase 2: Process ItemBlocks for a single backup in multiple threads
+
+#### New input field for number of ItemBlock workers
+
+The velero installer and server CLIs will get a new input field `itemBlockWorkerCount`, which will be passed along to the `backupReconciler`.
+The `backupReconciler` struct will also have this new field added. 
+
+#### Worker pool for item block processing
+
+A new type, `ItemBlockWorker` will be added which will manage a pool of worker goroutines which will process item blocks, a shared input channel for passing blocks to workers, and a WaitGroup to shut down cleanly when the reconciler exits.
+```go
+type ItemBlockWorkerPool struct {
+    itemBlockChannel chan ItemBlockInput
+    wg               *sync.WaitGroup
+    logger           logrus.FieldLogger
+}
+
+type ItemBlockInput struct {
+    itemBlock  *BackupItemBlock
+    returnChan chan ItemBlockReturn
+}
+
+type ItemBlockReturn struct {
+    itemBlock *BackupItemBlock
+    resources []schema.GroupResource
+    err       error
+}
+
+func (*p ItemBlockWorkerPool) getInputChannel() chan ItemBlockInput
+func StartItemBlockWorkerPool(context context.Context, workers int, logger logrus.FieldLogger) ItemBlockWorkerPool
+func processItemBlockWorker(context context.Context, itemBlockChannel chan ItemBlockInput, logger logrus.FieldLogger, wg *sync.WaitGroup)
+```
+
+The worker pool will be started by calling `StartItemBlockWorkerPool` in `NewBackupReconciler()`, passing in the worker count and reconciler context.
+`backupreconciler.prepareBackupRequest` will also add the input channel to the `backupRequest` so that it will be available during backup processing.
+The func `StartItemBlockWorkerPool` will create the `ItemBlockWorkerPool` with a shared buffered input channel (fixed buffer size) and start `workers` gororoutines which will each call `processItemBlockWorker`.
+The `processItemBlockWorker` func (run by the worker goroutines) will read from `itemBlockChannel`, call `BackupItemBlock` on the retrieved `ItemBlock`, and then send the return value to the retrieved `returnChan`, and then process the next block.
+
+#### Modify ItemBlock processing loop to send ItemBlocks to the worker pool rather than backing them up directly
+
+The ItemBlock processing loop implemented in Phase 1 will be modified to send each newly-created ItemBlock to the shared channel rather than calling `BackupItemBlock` inline, using a WaitGroup to manage in-process items. A separate goroutine will be created to process returns for this backup. After completion of the ItemBlock processing loop, velero will use the WaitGroup to wait for all ItemBlock processing to complete before moving forward.
+
+A simplified example of what this response goroutine might look like:
+```go
+    // omitting cancel handling, context, etc
+    ret := make(chan ItemBlockReturn)
+    wg := &sync.WaitGroup{}
+    // Handle returns
+    go func() {
+        for {
+            select {
+            case response := <-ret: // process each BackupItemBlock response
+                func() {
+                    defer wg.Done()
+                    responses = append(responses, response)
+                }()
+            case <-ctx.Done():
+                return
+            }
+        }
+    }()
+    // Simplified illustration, looping over and assumed already-determined ItemBlock list
+    for _, itemBlock := range itemBlocks {
+        wg.Add(1)
+        inputChan <- ItemBlockInput{itemBlock: itemBlock, returnChan: ret}
+    }
+    done := make(chan struct{})
+    go func() {
+        defer close(done)
+        wg.Wait()
+    }()
+    // Wait for all the ItemBlocks to be processed
+    select {
+    case <-done:
+        logger.Info("done processing ItemBlocks")
+    }
+    // responses from BackupItemBlock calls are in responses
+```
+
+When processing the responses, the main thing is to set `backedUpGroupResources[item.groupResource]=true` for each GR returned, which will give the same result as the current implementation calling items one-by-one and setting that field as needed.
+
+The ItemBlock processing loop described above will be split into two separate iterations. For the first iteration, velero will only process those items at the beginning of the loop identified as `orderedResources` -- when the groups generated from these resources are passed to the worker channel, velero will wait for the response before moving on to the next ItemBlock.
+This is to ensure that the ordered resources are processed in the required order. Once the last ordered resource is processed, the remaining ItemBlocks will be processed and sent to the worker channel without waiting for a response, in order to allow these ItemBlocks to be processed in parallel.
+The reason we must execute `ItemBlocks` with ordered resources first (and one at a time) is that this is a list of resources identified by the user as resources which must be backed up first, and in a particular order.
+
+#### Synchronize access to the BackedUpItems map
+
+Velero uses a map of BackedUpItems to track which items have already been backed up. This prevents velero from attempting to back up an item more than once, as well as guarding against creating infinite loops due to circular dependencies in the additional items returns. Since velero will now be accessing this map from the parallel goroutines, access to the map must be synchronized with mutexes.
+
+### Backup Finalize phase
+
+The finalize phase will not be affected by the ItemBlock design, since this is just updating resources after async operations are completed on the items and there is no need to run these updates in parallel.
+
+## Alternatives considered
+
+### BackpuItemAction v3 API
+
+Instead of adding a new  `ItemBlockAction` plugin type, we could add a `GetAdditionalItems` method to BackupItemAction.
+This was rejected because the new plugin type provides a cleaner interface, and keeps the function of grouping related items separate from the function of modifying item content for the backup.
+
+### Per-backup worker pool
+
+The current design makes use of a permanent worker pool, started at backup controller startup time. With this design, when we follow on with running multiple backups in parallel, the same set of workers will take ItemBlock inputs from more than one backup. Another approach that was initially considered was a temporary worker pool, created while processing a backup, and deleted upon backup completion. 
+
+#### User-visible API differences between the two approaches
+
+The main user-visible difference here is in the configuration API. For the permanent worker approach, the worker count represents the total worker count for all backups. The concurrent backup count represents the number of backups running at the same time. At any given time, though, the maximum number of worker threads backing up items concurrently is equal to the worker count. If worker count is 15 and the concurrent backup count is 3, then there will be, at most, 15 items being processed at the same time, split among up to three running backups.
+
+For the per-backup worker approach, the worker count represents the worker count for each backup. The concurrent backup count, as before, represents the number of backups running at the same time. If worker count is 15 and the concurrent backup count is 3, then there will be, at most, 45 items being processed at the same time, up to 15 for each of up to three running backups.
+#### Comparison of the two approaches
+
+- Permanent worker pool advantages:
+  - This is the more commonly-followed Kubernetes pattern. It's generally better to follow standard practices, unless there are genuine reasons for the use case to go in a different way.
+  - It's easier for users to understand the maximum number of concurrent items processed, which will have performance impact and impact on the resource requirements for the Velero pod. Users will not have to multiply the config numbers in their heads when working out how many total workers are present.
+  - It will give us more flexibility for future enhancements around concurrent backups. One possible use case: backup priority. Maybe a user wants scheduled backups to have a lower priority than user-generated backups, since a user is sitting there waiting for completion -- a shared worker pool could react to the priority by taking ItemBlocks for the higher priority backup first, which would allow a large lower-priority backup's items to be preempted by a higher-priority backup's items without needing to explicitly stop the main controller flow for that backup.
+- Per-backup worker pool advantages:
+  - Lower memory consumption than permanent worker pool, but the total memory used by a worker blocked on input will be pretty low, so if we're talking only 10-20 workers, the impact will be minimal.
+
+## Compatibility
+
+### Example IBA implementation for BIA plugins which return additional items
+
+Included below is an example of what might be required for a BIA  plugin which returns additional items.
+The code is taken from the internal velero `pod_action.go` which identifies the items required for a given pod.
+
+In this particular case, the only function of pod_action is to return additional items, so we can really just convert this plugin to an IBA plugin. If there were other actions, such as modifying the pod content on backup, then we would still need the pod action, and the related items vs. content manipulation functions would need to be separated.
+
+```go
+// PodAction implements ItemBlockAction.
+type PodAction struct {
+	log logrus.FieldLogger
+}
+
+// NewPodAction creates a new ItemAction for pods.
+func NewPodAction(logger logrus.FieldLogger) *PodAction {
+	return &PodAction{log: logger}
+}
+
+// AppliesTo returns a ResourceSelector that applies only to pods.
+func (a *PodAction) AppliesTo() (velero.ResourceSelector, error) {
+	return velero.ResourceSelector{
+		IncludedResources: []string{"pods"},
+	}, nil
+}
+
+// GetRelatedItems scans the pod's spec.volumes for persistentVolumeClaim volumes and returns a
+// ResourceIdentifier list containing references to all of the persistentVolumeClaim volumes used by
+// the pod. This ensures that when a pod is backed up, all referenced PVCs are backed up too.
+func (a *PodAction) GetRelatedItems(item runtime.Unstructured, backup *v1.Backup) (runtime.Unstructured, []velero.ResourceIdentifier, error) {
+	pod := new(corev1api.Pod)
+	if err := runtime.DefaultUnstructuredConverter.FromUnstructured(item.UnstructuredContent(), pod); err != nil {
+		return nil, errors.WithStack(err)
+	}
+
+	var relatedItems []velero.ResourceIdentifier
+	if pod.Spec.PriorityClassName != "" {
+		a.log.Infof("Adding priorityclass %s to relatedItems", pod.Spec.PriorityClassName)
+		relatedItems = append(relatedItems, velero.ResourceIdentifier{
+			GroupResource: kuberesource.PriorityClasses,
+			Name:          pod.Spec.PriorityClassName,
+		})
+	}
+
+	if len(pod.Spec.Volumes) == 0 {
+		a.log.Info("pod has no volumes")
+		return relatedItems, nil
+	}
+
+	for _, volume := range pod.Spec.Volumes {
+		if volume.PersistentVolumeClaim != nil && volume.PersistentVolumeClaim.ClaimName != "" {
+			a.log.Infof("Adding pvc %s to relatedItems", volume.PersistentVolumeClaim.ClaimName)
+
+			relatedItems = append(relatedItems, velero.ResourceIdentifier{
+				GroupResource: kuberesource.PersistentVolumeClaims,
+				Namespace:     pod.Namespace,
+				Name:          volume.PersistentVolumeClaim.ClaimName,
+			})
+		}
+	}
+
+	return relatedItems, nil
+}
+
+// API call
+func (a *PodAction) Name() string {
+	return "PodAction"
+}
+
+```
+
+
+## Implementation
+Phase 1 and Phase 2 could be implemented within the same Velero release cycle, but they need not be.
+Phase 1 is expected to be implemented in Velero 1.15.
+Phase 2 is expected to be implemented in Velero 1.16.

design/Implemented/backup-pvc-config.md

@@ -0,0 +1,94 @@
+# Backup PVC Configuration Design
+
+## Glossary & Abbreviation
+
+**Velero Generic Data Path (VGDP)**: VGDP is the collective modules that is introduced in [Unified Repository design][1]. Velero uses these modules to finish data transfer for various purposes (i.e., PodVolume backup/restore, Volume Snapshot Data Movement). VGDP modules include uploaders and the backup repository.  
+
+**Exposer**: Exposer is a module that is introduced in [Volume Snapshot Data Movement Design][2]. Velero uses this module to expose the volume snapshots to Velero node-agent pods or node-agent associated pods so as to complete the data movement from the snapshots.  
+
+**backupPVC**: The intermediate PVC created by the exposer for VGDP to access data from, see [Volume Snapshot Data Movement Design][2] for more details.  
+
+**backupPod**: The pod consumes the backupPVC so that VGDP could access data from the backupPVC, see [Volume Snapshot Data Movement Design][2] for more details.  
+
+**sourcePVC**: The PVC to be backed up, see [Volume Snapshot Data Movement Design][2] for more details. 
+
+## Background
+
+As elaberated in [Volume Snapshot Data Movement Design][2], a backupPVC may be created by the Exposer and the VGDP reads data from the backupPVC.  
+In some scenarios, users may need to configure some advanced settings of the backupPVC so that the data movement could work in best performance in their environments. Specifically:  
+- For some storage providers, when creating a read-only volume from a snapshot, it is very fast; whereas, if a writable volume is created from the snapshot, they need to clone the entire disk data, which is time consuming. If the backupPVC's `accessModes` is set as `ReadOnlyMany`, the volume driver is able to tell the storage to create a read-only volume, which may dramatically shorten the snapshot expose time. On the other hand,  `ReadOnlyMany` is not supported by all volumes. Therefore, users should be allowed to configure the `accessModes` for the backupPVC.  
+- Some storage providers create one or more replicas when creating a volume, the number of replicas is defined in the storage class. However, it doesn't make any sense to keep replicas when an intermediate volume used by the backup. Therefore, users should be allowed to configure another storage class specifically used by the backupPVC.  
+
+## Goals
+
+- Create a mechanism for users to specify various configurations for backupPVC    
+
+## Non-Goals
+
+## Solution
+
+We will use the ConfigMap specified by `velero node-agent` CLI's parameter `--node-agent-configmap` to host the backupPVC configurations.
+This configMap is not created by Velero, users should create it manually on demand. The configMap should be in the same namespace where Velero is installed. If multiple Velero instances are installed in different namespaces, there should be one configMap in each namespace which applies to node-agent in that namespace only.  
+Node-agent server checks these configurations at startup time and use it to initiate the related Exposer modules. Therefore, users could edit this configMap any time, but in order to make the changes effective, node-agent server needs to be restarted.  
+Inside the ConfigMap we will add one new kind of configuration as the data in the configMap, the name is ```backupPVC```.  
+Users may want to set different backupPVC configurations for different volumes, therefore, we define the configurations as a map and allow users to specific configurations by storage class. Specifically, the key of the map element is the storage class name used by the sourcePVC and the value is the set of configurations for the backupPVC created for the sourcePVC.   
+
+The data structure is as below:
+```go
+type Configs struct {
+	// LoadConcurrency is the config for data path load concurrency per node.
+	LoadConcurrency *LoadConcurrency `json:"loadConcurrency,omitempty"`
+
+	// LoadAffinity is the config for data path load affinity.
+	LoadAffinity []*LoadAffinity `json:"loadAffinity,omitempty"`
+
+	// BackupPVC is the config for backupPVC of snapshot data movement.
+	BackupPVC map[string]BackupPVC `json:"backupPVC,omitempty"`
+}
+
+type BackupPVC struct {
+	// StorageClass is the name of storage class to be used by the backupPVC.
+	StorageClass string `json:"storageClass,omitempty"`
+
+	// ReadOnly sets the backupPVC's access mode as read only.
+	ReadOnly bool `json:"readOnly,omitempty"`
+}
+```  
+
+### Sample
+A sample of the ConfigMap is as below:
+```json
+{
+    "backupPVC": {
+        "storage-class-1": {
+            "storageClass": "snapshot-storage-class",
+            "readOnly": true
+        },
+        "storage-class-2": {
+            "storageClass": "snapshot-storage-class"
+        },
+        "storage-class-3": {
+            "readOnly": true
+        }        
+    }
+}
+```
+
+To create the configMap, users need to save something like the above sample to a json file and then run below command:
+```
+kubectl create cm <ConfigMap name> -n velero --from-file=<json file name>
+``` 
+
+### Implementation
+The `backupPVC` is passed to the exposer and the exposer sets the related specification and create the backupPVC.  
+If `backupPVC.storageClass` doesn't exist or set as empty, the sourcePVC's storage class will be used.  
+If `backupPVC.readOnly` is set to true, `ReadOnlyMany` will be the only value set to the backupPVC's `accessModes`, otherwise, `ReadWriteOnce` is used.  
+
+Once `backupPVC.storageClass` is set, users must make sure that the specified storage class exists in the cluster and can be used the the backupPVC, otherwise, the corresponding DataUpload CR will stay in `Accepted` phase until the prepare timeout (by default 30min).   
+Once `backupPVC.readOnly` is set to true, users must make sure that the storage supports to create a `ReadOnlyMany` PVC from a snapshot, otherwise, the corresponding DataUpload CR will stay in `Accepted` phase until the prepare timeout (by default 30min).  
+
+Once above problems happen, the DataUpload CR is cancelled after prepare timeout and the backupPVC and backupPod will be deleted, so there is no way to tell the cause is one of the above problems or others.  
+To help the troubleshooting, we can add some diagnostic mechanism to discover the status of the backupPod before deleting it as a result of the prepare timeout.  
+
+[1]: unified-repo-and-kopia-integration/unified-repo-and-kopia-integration.md
+[2]: volume-snapshot-data-movement/volume-snapshot-data-movement.md

design/Implemented/backup-repo-cache-volume.md

@@ -0,0 +1,231 @@
+# Backup Repository Cache Volume Design
+
+## Glossary & Abbreviation
+
+**Backup Storage**: The storage to store the backup data. Check [Unified Repository design][1] for details.  
+**Backup Repository**: Backup repository is layered between BR data movers and Backup Storage to provide BR related features that is introduced in [Unified Repository design][1].  
+**Velero Generic Data Path (VGDP)**: VGDP is the collective of modules that is introduced in [Unified Repository design][1]. Velero uses these modules to finish data transfer for various purposes (i.e., PodVolume backup/restore, Volume Snapshot Data Movement). VGDP modules include uploaders and the backup repository.  
+**Data Mover Pods**: Intermediate pods which hold VGDP and complete the data transfer. See [VGDP Micro Service for Volume Snapshot Data Movement][2] and [VGDP Micro Service For fs-backup][3] for details.  
+**Repository Maintenance Pods**: Pods for [Repository Maintenance Jobs][4], which holds VGDP to run repository maintenance.    
+
+## Background
+
+According to the [Unified Repository design][1] Velero uses selectable backup repositories for various backup/restore methods, i.e., fs-backup, volume snapshot data movement, etc. Some backup repositories may need to cache data on the client side for various repository operation, so as to accelerate the execution.  
+In the existing [Backup Repository Configuration][5], we allow users to configure the cache data size (`cacheLimitMB`). However, the cache data is still stored in the root file system of data mover pods/repository maintenance pods, so stored in the root file system of the node. This is not good enough, reasons:  
+- In many distributions, the node's system disk size is predefined, non configurable and limit, e.g., the system disk size may be 20G or less
+- Velero supports concurrent data movements in each node. The cache in each of the concurrent data mover pods could quickly run out of the system disk and cause problems like pod eviction, failure of pod creation, degradation of Kubernetes QoS, etc.  
+
+We need to allow users to prepare a dedicated location, e.g., a dedictated volume, for the cache.  
+Not all backup repositories or not all backup repository operations require cache, we need to define the details when and how the cache is used.  
+
+## Goals
+
+- Create a mechanism for users to configure cache volumes for various pods running VGDP
+- Design the workflow to assign the cache volume pod path to backup repositories
+- Describe when and how the cache volume is used 
+
+## Non-Goals
+
+- The solution is based on [Unified Repository design][1], [VGDP Micro Service for Volume Snapshot Data Movement][2] and [VGDP Micro Service For fs-backup][3], legacy data paths are not supported. E.g., when a pod volume restore (PVR) runs with legacy Restic path, if any data is cached, the cache still resides in the root file system.  
+
+## Solution
+
+### Cache Data
+
+Varying on backup repositoires, cache data may include payload data or repository metadata, e.g., indexes to the payload data chunks.  
+
+Payload data is highly related to the backup data, and normally take the majority of the repository data as well as the cache data.
+
+Repository metadata is related to the backup repository's chunking algorithm, data chunk mapping method, etc, and so the size is not proportional to the backup data size.  
+On the other hand for some backup repository, in extreme cases, the repository metadata may be significantly large. E.g., Kopia's indexes are per chunks, if there are huge number of small files in the repository, Kopia's index data may be in the same level of or even larger than the payload data.    
+However, in the cases that repository metadata data become the majority, other bottlenecks may emerge and concurrency of data movers may be significantly constrained, so the requirement to cache volumes may go away.  
+
+Therefore, for now we only consider the cache volume requirement for payload data, and leave the consideration for metadata as a future enhancement.  
+
+### Scenarios
+
+Backup repository cache varies on backup repositories and backup repository operation during VGDP runs. Below are the scenarios when VGDP runs:
+- Data Upload for Backup: this is the process to upload/write the backup data into the backup repository, e.g., DataUpload or PodVolumeBackup. The pieces of data is almost directly written to the repository, sometimes with a small group staying shortly in the local place. That is to say, there should not be large scale data cached for this scenario, so we don't prepare dedicated cache for this scenario.
+- Repository Maintenance: Repository maintenance most often visits the backup repository's metadata and sometimes it needs to visit the file system directories from the backed up data. On the other hand, it is not practical to run concurrent maintenance jobs in one node. So the cache data is neither large nor affect the root file system too much. Therefore, we don't need to prepare dedicated cache for this scenario.
+- Data Download for Restore: this is the process to download/read the backup data from the backup repository during restore, e.g., DataDownload or PodVolumeRestore. For backup repositories for which data are stored in remote backup storages (e.g., Kopia repository stores data in remote object stores), large scale of data are cached locally to accerlerate the restore. Therefore, we need dedicate cache volumes for this scenario.  
+- Backup Deletion: During this scenario, backup repository is connected, metadata is enumerated to find the repository snapshot representing the backup data. That is to say, only metadata is cached if any. Therefore, dedicated cache volumes are not required in this scenario.
+
+The above analyses are based on the common behavior of backup repositories and they are not considering the case that backup repository metadata takes majority or siginficant proportion of the cache data.   
+As a conclusion of the analyses, we will create dedicated cache volumes for restore scenarios.  
+For other scenarios, we can add them regarded to the future changes/requirements. The mechanism to expose and connect the cache volumes should work for all scenarios. E.g., if we need to consider the backup repository metadata case, we may need cache volumes for backup and repository maintenance as well, then we can just reuse the same cache volume provision and connection mechanism to backup and repository maintenance scenarios.   
+
+### Cache Data and Lifecycle
+
+If available, one cache volume is dedicately assigned to one data mover pod. That is, the cached data is destroyed when the data mover pod completes. Then the backup repository instance also closes.    
+Cache data are fully managed by the specific backup repository. So the backup repository may also have its own way to GC the cache data.  
+That is to say, cache data GC may be launched by the backup repository instance during the running of the data mover pod; then the left data are automatically destroyed when the data mover pod and the cache PVC are destroyed (cache PVC's `reclaimPolicy` is always `Deleted`, so once the cache PVC is destroyed, the volume will also be destroyed). So no specially logics are needed for cache data GC.  
+
+### Data Size
+
+Cache volumes take storage space and cluster resources (PVC, PV), therefore, cache volumes should be created only when necessary and the volumes should be with reasonable size based on the cache data size:  
+- It is not a good bargain to have cache volumes for small backups, small backups will use resident cache location (the cache location in the root file system)
+- The cache data size has a limit, the existing `cacheLimitMB` is used for this purpose. E.g., it could be set as 1024 for a 1TB backup, which means 1GB of data is cached and the old cache data exceeding this size will be cleared. Therefore, it is meaningless to set the cache volume size much larger than `cacheLimitMB`
+
+### Cache Volume Size
+
+The cache volume size is calculated from below factors (for Restore scenarios):  
+- **Limit**: The limit of the cache data, that is represented by `cacheLimitMB`, the default value is 5GB
+- **backupSize**: The size of the backup as a reference to evaluate whether to create a cache volume. It doesn't mean the backup data really decides the cache data all the time, it is just a reference to evaluate the scale of the backup, small scale backups may need small cache data. Sometimes, backupSize is not irrelevant to the size of cache data, in this case, ResidentThreshold should not be set, Limit will be used directly. It is unlikely that backupSize is unavailable, but once that happens, ResidentThreshold is ignored, Limit will be used directly.  
+- **ResidentThreshold**: The minimum backup size that a cache volume is created
+- **InflationPercentage**: Considering the overhead of the file system and the possible delay of the cache cleanup, there should be an inflation for the final volume size vs. the logical size, otherwise, the cache volume may be overrun. This inflation percentage is hardcoded, e.g., 20%. 
+
+A formula is as below:  
+```
+cacheVolumeSize = ((backupSize != 0 ? (backupSize > residentThreshold ? limit : 0) : limit) * (100 + inflationPercentage)) / 100
+```
+Finally, the `cacheVolumeSize` will be rounded up to GiB considering the UX friendliness, storage friendliness and management friendliness.    
+
+### PVC/PV
+
+The PVC for a cache volume is created in Velero namespace and a storage class is required for the cache PVC. The PVC's accessMode is `ReadWriteOnce` and volumeMode is `FileSystem`, so the storage class provided should support this specification. Otherwise, if the storageclass doesn't support either of the specifications, the data mover pod may be hang in `Pending` state until a timeout setting with the data movement (e.g. `prepareTimeout`) and the data movement will finally fail.  
+It is not expected that the cache volume is retained after data mover pod is deleted, so the `reclaimPolicy` for the storageclass must be `Delete`.  
+
+To detect the problems in the storageclass and fail earlier, a validation is applied to the storageclass and once the validation fails, the cache configuration will be ignored, so the data mover pod will be created without a cache volume.  
+
+### Cache Volume Configurations
+
+Below configurations are introduced:
+- **residentThresholdMB**: the minimum data size(in MB) to be processed (if available) that a cache volume is created
+- **cacheStorageClass**: the name of the storage class to provision the cache PVC
+
+Not like `cacheLimitMB` which is set to and affect the backup repository, the above two configurations are actually data mover configurations of how to create cache volumes to data mover pods; and the two configurations don't need to be per backup repository. So we add them to the node-agent Configuration.  
+
+### Sample
+
+Below are some examples of the node-agent configMap with the configurations:
+
+Sample-1:  
+```json
+{
+    "cacheVolume": {
+        "storageClass": "sc-1",
+        "residentThresholdMB": 1024        
+    }
+}
+```
+
+Sample-2:  
+```json
+{
+    "cacheVolume": {
+        "storageClass": "sc-1",       
+    }
+}
+```
+
+Sample-3:  
+```json
+{
+    "cacheVolume": {
+        "residentThresholdMB": 1024        
+    }
+}
+```
+
+**sample-1**: This is a valid configuration. Restores with backup data size larger than 1G will be assigned a cache volume using storage class `sc-1`.  
+**sample-2**: This is a valid configuration. Data mover pods are always assigned a cache volume using storage class `sc-1`.  
+**sample-3**: This is not a valid configuration because the storage class is absent. Velero gives up creating a cache volume.   
+
+To create the configMap, users need to save something like the above sample to a json file and then run below command:
+```
+kubectl create cm <ConfigMap name> -n velero --from-file=<json file name>
+```
+
+The cache volume configurations will be visited by node-agent server, so they also need to specify the `--node-agent-configmap` to the `velero node-agent` parameters.  
+
+## Detailed Design
+
+### Backup and Restore
+
+The restore needs to know the backup size so as to calculate the cache volume size, some new fields are added to the DataDownload and PodVolumeRestore CRDs.  
+
+`snapshotSize` field is also added to DataDownload and PodVolumeRestore's `spec`:
+```yaml
+          spec:
+              snapshotID:
+                description: SnapshotID is the ID of the Velero backup snapshot to
+                  be restored from.
+                type: string
+              snapshotSize:
+                description: SnapshotSize is the logical size of the snapshot.
+                format: int64
+                type: integer
+```
+
+`snapshotSize` represents the total size of the backup; during restore, the value is transferred from DataUpload/PodVolumeBackup's `Status.Progress.TotalBytes` to DataDownload/PodVolumeRestore.    
+
+It is unlikely that `Status.Progress.TotalBytes` from DataUpload/PodVolumeBackup is unavailable, but once it happens, according to the above formula, `residentThresholdMB` is ignored, cache volume size is calculated directly from cache limit for the corresponding backup repository.  
+
+### Exposer
+
+Cache volume configurations are retrieved by node-agent and passed through DataDownload/PodVolumeRestore to GenericRestore exposer/PodVolume exposer.  
+The exposers are responsible to calculate cache volume size, create cache PVCs and mount them to the restorePods.  
+If the calculated cache volume size is 0, or any of the critical parameters is missing (e.g., cache volume storage class), the exposers ignore the cache volume configuration and continue with creating restorePods without cache volumes, so no impact to the result of the restore.  
+
+Exposers mount the cache volume to a predefined directory and pass the directory to the data mover pods through the `cache-volume-path` parameter.  
+
+Below data structure is added to the exposers' expose parameters:  
+
+```go
+type GenericRestoreExposeParam struct {
+	// RestoreSize specifies the data size for the volume to be restored
+	RestoreSize int64
+
+	// CacheVolume specifies the info for cache volumes
+	CacheVolume *CacheVolumeInfo
+}
+
+type PodVolumeExposeParam struct {
+	// RestoreSize specifies the data size for the volume to be restored
+	RestoreSize int64
+
+	// CacheVolume specifies the info for cache volumes
+	CacheVolume *repocache.CacheConfigs
+}
+
+type CacheConfigs struct {
+	// StorageClass specifies the storage class for cache volumes
+	StorageClass string
+
+	// Limit specifies the maximum size of the cache data
+	Limit int64
+
+	// ResidentThreshold specifies the minimum size of the cache data to create a cache volume
+	ResidentThreshold int64
+}
+```
+
+### Data Mover Pods
+
+Data mover pods retrieve the cache volume directory from `cache-volume-path` parameter and pass it to Unified Repository.  
+If the directory is empty, Unified Repository uses the resident location for data cache, that is, the root file system.  
+
+### Kopia Repository
+
+Kopia repository supports cache directory configuration for both metadata and data. The existing `SetupConnectOptions` is modified to customize the `CacheDirectory`:  
+
+```go
+func SetupConnectOptions(ctx context.Context, repoOptions udmrepo.RepoOptions) repo.ConnectOptions {
+    ...
+
+	return repo.ConnectOptions{
+		CachingOptions: content.CachingOptions{
+			CacheDirectory: cacheDir,
+			...
+		},
+		...
+	}
+}
+```  
+
+
+[1]: Implemented/unified-repo-and-kopia-integration/unified-repo-and-kopia-integration.md
+[2]: Implemented/vgdp-micro-service/vgdp-micro-service.md
+[3]: Implemented/vgdp-micro-service-for-fs-backup/vgdp-micro-service-for-fs-backup.md
+[4]: Implemented/repo_maintenance_job_config.md
+[5]: Implemented/backup-repo-config.md

design/Implemented/backup-repo-config.md

@@ -0,0 +1,123 @@
+# Backup Repository Configuration Design
+
+## Glossary & Abbreviation
+
+**Backup Storage**: The storage to store the backup data. Check [Unified Repository design][1] for details.  
+**Backup Repository**: Backup repository is layered between BR data movers and Backup Storage to provide BR related features that is introduced in [Unified Repository design][1].    
+
+## Background
+
+According to the [Unified Repository design][1] Velero uses selectable backup repositories for various backup/restore methods, i.e., fs-backup, volume snapshot data movement, etc. To achieve the best performance, backup repositories may need to be configured according to the running environments.  
+For example, if there are sufficient CPU and memory resources in the environment, users may enable compression feature provided by the backup repository, so as to achieve the best backup throughput.  
+As another example, if the local disk space is not sufficient, users may want to constraint the backup repository's cache size, so as to prevent the repository from running out of the disk space.  
+Therefore, it is worthy to allow users to configure some essential parameters of the backup repsoitories, and the configuration may vary from backup repositories.  
+
+## Goals
+
+- Create a mechanism for users to specify configurations for backup repositories  
+
+## Non-Goals
+
+## Solution
+
+### BackupRepository CRD
+
+After a backup repository is initialized, a BackupRepository CR is created to represent the instance of the backup repository. The BackupRepository's spec is a core parameter used by Unified Repo modules when interactive with the backup repsoitory. Therefore, we can add the configurations into the BackupRepository CR called ```repositoryConfig```.  
+The configurations may be different varying from backup repositories, therefore, we will not define each of the configurations explicitly. Instead, we add a map in the BackupRepository's spec to take any configuration to be set to the backup repository.  
+
+During various operations to the backup repository, the Unified Repo modules will retrieve from the map for the specific configuration that is required at that time. So even though it is specified, a configuration may not be visited/hornored if the operations don't require it for the specific backup repository, this won't bring any issue. When and how a configuration is hornored is decided by the configuration itself and should be clarified in the configuration's specification.  
+
+Below is the new BackupRepository's spec after adding the configuration map:  
+```yaml
+          spec:
+            description: BackupRepositorySpec is the specification for a BackupRepository.
+            properties:
+              backupStorageLocation:
+                description: |-
+                  BackupStorageLocation is the name of the BackupStorageLocation
+                  that should contain this repository.
+                type: string
+              maintenanceFrequency:
+                description: MaintenanceFrequency is how often maintenance should
+                  be run.
+                type: string
+              repositoryConfig:
+                additionalProperties:
+                  type: string
+                description: RepositoryConfig contains configurations for the specific
+                  repository.
+                type: object
+              repositoryType:
+                description: RepositoryType indicates the type of the backend repository
+                enum:
+                - kopia
+                - restic
+                - ""
+                type: string
+              resticIdentifier:
+                description: |-
+                  ResticIdentifier is the full restic-compatible string for identifying
+                  this repository.
+                type: string
+              volumeNamespace:
+                description: |-
+                  VolumeNamespace is the namespace this backup repository contains
+                  pod volume backups for.
+                type: string
+            required:
+            - backupStorageLocation
+            - maintenanceFrequency
+            - resticIdentifier
+            - volumeNamespace
+            type: object
+```            
+
+### BackupRepository configMap
+
+The BackupRepository CR is not created explicitly by a Velero CLI, but created as part of the backup/restore/maintenance operation if the CR doesn't exist. As a result, users don't have any way to specify the configurations before the BackupRepository CR is created.  
+Therefore, a BackupRepository configMap is introduced as a template of the configurations to be applied to the backup repository CR.  
+When the backup repository CR is created by the BackupRepository controller, the configurations in the configMap are copied to the ```repositoryConfig``` field.   
+For an existing BackupRepository CR, the configMap is never visited, if users want to modify the configuration value, they should directly edit the BackupRepository CR.  
+
+The BackupRepository configMap is created by users in velero installation namespace. The configMap name must be specified in the velero server parameter ```--backup-repository-configmap```, otherwise, it won't effect.  
+If the configMap name is specified but the configMap doesn't exist by the time of a backup repository is created, the configMap name is ignored.  
+For any reason, if the configMap doesn't effect, nothing is specified to the backup repository CR, so the Unified Repo modules use the hard-coded values to configure the backup repository.  
+
+The BackupRepository configMap supports backup repository type specific configurations, even though users can only specify one configMap.  
+So in the configMap struct, multiple entries are supported, indexed by the backup repository type. During the backup repository creation, the configMap is searched by the repository type.  
+
+### Configurations
+
+With the above mechanisms, any kind of configuration could be added. Here list the configurations defined at present:  
+```cacheLimitMB```: specifies the size limit(in MB) for the local data cache. The more data is cached locally, the less data may be downloaded from the backup storage, so the better performance may be achieved. Practically, users can specify any size that is smaller than the free space so that the disk space won't run out. This parameter is for each repository connection, that is, users could change it before connecting to the repository. If a backup repository doesn't use local cache, this parameter will be ignored. For Kopia repository, this parameter is supported.  
+```enableCompression```: specifies to enable/disable compression for a backup repsotiory. Most of the backup repositories support the data compression feature, if it is not supported by a backup repository, this parameter is ignored. Most of the backup repositories support to dynamically enable/disable compression, so this parameter is defined to be used whenever creating a write connection to the backup repository, if the dynamically changing is not supported, this parameter will be hornored only when initializing the backup repository. For Kopia repository, this parameter is supported and can be dynamically modified.  
+
+### Sample
+Below is an example of the BackupRepository configMap with the configurations:     
+```yaml
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: <config-name>
+  namespace: velero
+data:
+  <repository-type-1>: |
+    {
+      "cacheLimitMB": 2048,
+      "enableCompression": true    
+    }
+  <repository-type-2>: |
+    {
+      "cacheLimitMB": 1,
+      "enableCompression": false    
+    }        
+```
+
+To create the configMap, users need to save something like the above sample to a file and then run below commands:  
+```
+kubectl apply -f <yaml file name>
+```  
+
+
+
+[1]: unified-repo-and-kopia-integration/unified-repo-and-kopia-integration.md

design/Implemented/bsl-certificate-support_design.md

@@ -0,0 +1,417 @@
+# Design for BSL Certificate Support Enhancement
+
+## Abstract
+
+This design document describes the enhancement of BackupStorageLocation (BSL) certificate management in Velero, introducing a Secret-based certificate reference mechanism (`caCertRef`) alongside the existing inline certificate field (`caCert`). This enhancement provides a more secure, Kubernetes-native approach to certificate management while enabling future CLI improvements for automatic certificate discovery.
+
+## Background
+
+Currently, Velero supports TLS certificate verification for object storage providers through an inline `caCert` field in the BSL specification. While functional, this approach has several limitations:
+
+- **Security**: Certificates are stored directly in the BSL YAML, potentially exposing sensitive data
+- **Management**: Certificate rotation requires updating the BSL resource itself
+- **CLI Usability**: Users must manually specify certificates when using CLI commands
+- **Size Limitations**: Large certificate bundles can make BSL resources unwieldy
+
+Issue #9097 and PR #8557 highlight the need for improved certificate management that addresses these concerns while maintaining backward compatibility.
+
+## Goals
+
+- Provide a secure, Secret-based certificate storage mechanism
+- Maintain full backward compatibility with existing BSL configurations
+- Enable future CLI enhancements for automatic certificate discovery
+- Simplify certificate rotation and management
+- Provide clear migration path for existing users
+
+## Non-Goals
+
+- Removing support for inline certificates immediately
+- Changing the behavior of existing BSL configurations
+- Implementing client-side certificate validation
+- Supporting certificates from ConfigMaps or other resource types
+
+## High-Level Design
+
+### API Changes
+
+#### New Field: CACertRef
+
+```go
+type ObjectStorageLocation struct {
+    // Existing field (now deprecated)
+    // +optional
+    // +kubebuilder:deprecatedversion:warning="caCert is deprecated, use caCertRef instead"
+    CACert []byte `json:"caCert,omitempty"`
+
+    // New field for Secret reference
+    // +optional
+    CACertRef *corev1api.SecretKeySelector `json:"caCertRef,omitempty"`
+}
+```
+
+The `SecretKeySelector` follows standard Kubernetes patterns:
+```go
+type SecretKeySelector struct {
+    // Name of the Secret
+    Name string `json:"name"`
+    // Key within the Secret
+    Key string `json:"key"`
+}
+```
+
+### Certificate Resolution Logic
+
+The system follows a priority-based resolution:
+
+1. If `caCertRef` is specified, retrieve certificate from the referenced Secret
+2. If `caCert` is specified (and `caCertRef` is not), use the inline certificate
+3. If neither is specified, no custom CA certificate is used
+
+### Validation
+
+BSL validation ensures mutual exclusivity:
+```go
+func (bsl *BackupStorageLocation) Validate() error {
+    if bsl.Spec.ObjectStorage != nil &&
+        bsl.Spec.ObjectStorage.CACert != nil &&
+        bsl.Spec.ObjectStorage.CACertRef != nil {
+        return errors.New("cannot specify both caCert and caCertRef in objectStorage")
+    }
+    return nil
+}
+```
+
+## Detailed Design
+
+### BSL Controller Changes
+
+The BSL controller incorporates validation during reconciliation:
+
+```go
+func (r *backupStorageLocationReconciler) Reconcile(req ctrl.Request) (ctrl.Result, error) {
+    // ... existing code ...
+    
+    // Validate BSL configuration
+    if err := location.Validate(); err != nil {
+        r.logger.WithError(err).Error("BSL validation failed")
+        return ctrl.Result{}, err
+    }
+    
+    // ... continue reconciliation ...
+}
+```
+
+### Repository Provider Integration
+
+All repository providers implement consistent certificate handling:
+
+```go
+func configureCACert(bsl *velerov1api.BackupStorageLocation, credGetter *credentials.CredentialGetter) ([]byte, error) {
+    if bsl.Spec.ObjectStorage == nil {
+        return nil, nil
+    }
+
+    // Prefer caCertRef (new method)
+    if bsl.Spec.ObjectStorage.CACertRef != nil {
+        certString, err := credGetter.FromSecret.Get(bsl.Spec.ObjectStorage.CACertRef)
+        if err != nil {
+            return nil, errors.Wrap(err, "error getting CA certificate from secret")
+        }
+        return []byte(certString), nil
+    }
+
+    // Fall back to caCert (deprecated)
+    if bsl.Spec.ObjectStorage.CACert != nil {
+        return bsl.Spec.ObjectStorage.CACert, nil
+    }
+
+    return nil, nil
+}
+```
+
+### CLI Certificate Discovery Integration
+
+#### Background: PR #8557 Implementation
+PR #8557 ("CLI automatically discovers and uses cacert from BSL") was merged in August 2025, introducing automatic CA certificate discovery from BackupStorageLocation for Velero CLI download operations. This eliminated the need for users to manually specify the `--cacert` flag when performing operations like `backup describe`, `backup download`, `backup logs`, and `restore logs`.
+
+#### Current Implementation (Post PR #8557)
+The CLI now automatically discovers certificates from BSL through the `pkg/cmd/util/cacert/bsl_cacert.go` module:
+
+```go
+// Current implementation only supports inline caCert
+func GetCACertFromBSL(ctx context.Context, client kbclient.Client, namespace, bslName string) (string, error) {
+    // ... fetch BSL ...
+    if bsl.Spec.ObjectStorage != nil && len(bsl.Spec.ObjectStorage.CACert) > 0 {
+        return string(bsl.Spec.ObjectStorage.CACert), nil
+    }
+    return "", nil
+}
+```
+
+#### Enhancement with caCertRef Support
+This design extends the existing CLI certificate discovery to support the new `caCertRef` field:
+
+```go
+// Enhanced implementation supporting both caCert and caCertRef
+func GetCACertFromBSL(ctx context.Context, client kbclient.Client, namespace, bslName string) (string, error) {
+    // ... fetch BSL ...
+
+    // Prefer caCertRef over inline caCert
+    if bsl.Spec.ObjectStorage.CACertRef != nil {
+        secret := &corev1api.Secret{}
+        key := types.NamespacedName{
+            Name:      bsl.Spec.ObjectStorage.CACertRef.Name,
+            Namespace: namespace,
+        }
+        if err := client.Get(ctx, key, secret); err != nil {
+            return "", errors.Wrap(err, "error getting certificate secret")
+        }
+
+        certData, ok := secret.Data[bsl.Spec.ObjectStorage.CACertRef.Key]
+        if !ok {
+            return "", errors.Errorf("key %s not found in secret",
+                bsl.Spec.ObjectStorage.CACertRef.Key)
+        }
+        return string(certData), nil
+    }
+
+    // Fall back to inline caCert (deprecated)
+    if bsl.Spec.ObjectStorage.CACert != nil {
+        return string(bsl.Spec.ObjectStorage.CACert), nil
+    }
+
+    return "", nil
+}
+```
+
+#### Certificate Resolution Priority
+
+The CLI follows this priority order for certificate resolution:
+
+1. **`--cacert` flag** - Manual override, highest priority
+2. **`caCertRef`** - Secret-based certificate (recommended)
+3. **`caCert`** - Inline certificate (deprecated)
+4. **System certificate pool** - Default fallback
+
+#### User Experience Improvements
+
+With both PR #8557 and this enhancement:
+
+```bash
+# Automatic discovery - works with both caCert and caCertRef
+velero backup describe my-backup
+velero backup download my-backup
+velero backup logs my-backup
+velero restore logs my-restore
+
+# Manual override still available
+velero backup describe my-backup --cacert /custom/ca.crt
+
+# Debug output shows certificate source
+velero backup download my-backup --log-level=debug
+# [DEBUG] Resolved CA certificate from BSL 'default' Secret 'storage-ca-cert' key 'ca-bundle.crt'
+```
+
+#### RBAC Considerations for CLI
+
+CLI users need read access to Secrets when using `caCertRef`:
+
+```yaml
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: velero-cli-user
+  namespace: velero
+rules:
+- apiGroups: ["velero.io"]
+  resources: ["backups", "restores", "backupstoragelocations"]
+  verbs: ["get", "list"]
+- apiGroups: [""]
+  resources: ["secrets"]
+  verbs: ["get"]
+  # Limited to secrets referenced by BSLs
+```
+
+### Migration Strategy
+
+#### Phase 1: Introduction (Current)
+- Add `caCertRef` field
+- Mark `caCert` as deprecated
+- Both fields supported, mutual exclusivity enforced
+
+#### Phase 2: Migration Period
+- Documentation and tools to help users migrate
+- Warning messages for `caCert` usage
+- CLI enhancements to leverage `caCertRef`
+
+#### Phase 3: Future Removal
+- Remove `caCert` field in major version update
+- Provide migration tool for automatic conversion
+
+## User Experience
+
+### Creating a BSL with Certificate Reference
+
+1. Create a Secret containing the CA certificate:
+```yaml
+apiVersion: v1
+kind: Secret
+metadata:
+  name: storage-ca-cert
+  namespace: velero
+type: Opaque
+data:
+  ca-bundle.crt: <base64-encoded-certificate>
+```
+
+2. Reference the Secret in BSL:
+```yaml
+apiVersion: velero.io/v1
+kind: BackupStorageLocation
+metadata:
+  name: default
+  namespace: velero
+spec:
+  provider: aws
+  objectStorage:
+    bucket: my-bucket
+    caCertRef:
+      name: storage-ca-cert
+      key: ca-bundle.crt
+```
+
+### Certificate Rotation
+
+With Secret-based certificates:
+```bash
+# Update the Secret with new certificate
+kubectl create secret generic storage-ca-cert \
+  --from-file=ca-bundle.crt=new-ca.crt \
+  --dry-run=client -o yaml | kubectl apply -f -
+
+# No BSL update required - changes take effect on next use
+```
+
+### CLI Usage Examples
+
+#### Immediate Benefits
+- No change required for existing workflows
+- Certificate validation errors include helpful context
+
+#### Future CLI Enhancements
+```bash
+# Automatic certificate discovery
+velero backup download my-backup
+
+# Manual override still available
+velero backup download my-backup --cacert /custom/ca.crt
+
+# Debug certificate resolution
+velero backup download my-backup --log-level=debug
+# [DEBUG] Resolved CA certificate from BSL 'default' Secret 'storage-ca-cert'
+```
+
+## Security Considerations
+
+### Advantages of Secret-based Storage
+
+1. **Encryption at Rest**: Secrets are encrypted in etcd
+2. **RBAC Control**: Fine-grained access control via Kubernetes RBAC
+3. **Audit Trail**: Secret access is auditable
+4. **Separation of Concerns**: Certificates separate from configuration
+
+### Required Permissions
+
+The Velero server requires additional RBAC permissions:
+```yaml
+- apiGroups: [""]
+  resources: ["secrets"]
+  verbs: ["get"]
+  # Scoped to secrets referenced by BSLs
+```
+
+## Compatibility
+
+### Backward Compatibility
+
+- Existing BSLs with `caCert` continue to function unchanged
+- No breaking changes to API
+- Gradual migration path
+
+### Forward Compatibility
+
+- Design allows for future enhancements:
+  - Multiple certificate support
+  - Certificate chain validation
+  - Automatic certificate discovery from cloud providers
+
+## Implementation Phases
+
+### Phase 1: Core Implementation ✓ (Current PR)
+- API changes with new `caCertRef` field
+- Controller validation
+- Repository provider updates
+- Basic testing
+
+### Phase 2: CLI Enhancement (Future)
+- Automatic certificate discovery in CLI
+- Enhanced error messages
+- Debug logging for certificate resolution
+
+### Phase 3: Migration Tools (Future)
+- Automated migration scripts
+- Validation tools
+- Documentation updates
+
+## Testing
+
+### Unit Tests
+- BSL validation logic
+- Certificate resolution in providers
+- Controller behavior
+
+### Integration Tests
+- End-to-end backup/restore with `caCertRef`
+- Certificate rotation scenarios
+- Migration from `caCert` to `caCertRef`
+
+### Manual Testing Scenarios
+1. Create BSL with `caCertRef`
+2. Perform backup/restore operations
+3. Rotate certificate in Secret
+4. Verify continued operation
+
+## Documentation
+
+### User Documentation
+- Migration guide from `caCert` to `caCertRef`
+- Examples for common cloud providers
+- Troubleshooting guide
+
+### API Documentation
+- Updated API reference
+- Deprecation notices
+- Field descriptions
+
+## Alternatives Considered
+
+### ConfigMap-based Storage
+- Pros: Similar to Secrets, simpler API
+- Cons: Not designed for sensitive data, no encryption at rest
+- Decision: Secrets are the Kubernetes-standard for sensitive data
+
+### External Certificate Management
+- Pros: Integration with cert-manager, etc.
+- Cons: Additional complexity, dependencies
+- Decision: Keep it simple, allow users to manage certificates as needed
+
+### Immediate Removal of Inline Certificates
+- Pros: Cleaner API, forces best practices
+- Cons: Breaking change, migration burden
+- Decision: Gradual deprecation respects existing users
+
+## Conclusion
+
+This design provides a secure, Kubernetes-native approach to certificate management in Velero while maintaining backward compatibility. It establishes the foundation for enhanced CLI functionality and improved user experience, addressing the concerns raised in issue #9097 and enabling the features proposed in PR #8557.
+
+The phased approach ensures smooth migration for existing users while delivering immediate security benefits for new deployments.

design/Implemented/clean_artifacts_in_csi_flow.md

@@ -0,0 +1,374 @@
+# Design to clean the artifacts generated in the CSI backup and restore workflows
+
+## Terminology
+
+* VSC: VolumeSnapshotContent
+* VS: VolumeSnapshot
+
+## Abstract
+* The design aims to delete the unnecessary VSs and VSCs generated during CSI backup and restore process. 
+* The design stop creating related VSCs during backup syncing.
+
+## Background
+In the current CSI backup and restore workflows, please notice the CSI B/R workflows means only using the CSI snapshots in the B/R, not including the CSI snapshot data movement workflows, some generated artifacts are kept after the backup or the restore process completion.
+
+Some of them are kept due to design, for example, the VolumeSnapshotContents generated during the backup are kept to make sure the backup deletion can clean the snapshots in the storage providers.
+
+Some of them are kept by accident, for example, after restore, two VolumeSnapshotContents are generated for the same VolumeSnapshot. One is from the backup content, and one is dynamically generated from the restore's VolumeSnapshot.
+
+The design aims to clean the unnecessary artifacts, and make the CSI B/R workflow more concise and reliable.
+
+## Goals
+- Clean the redundant VSC generated during CSI backup and restore.
+- Remove the VSCs in the backup sync process.
+
+## Non Goals
+- There were some discussion about whether Velero backup should include VSs and VSCs not generated in during the backup. By far, the conclusion is not including them is a better option. Although that is a useful enhancement, that is not included this design.
+- Delete all the CSI-related metadata files in the BSL is not the aim of this design. 
+
+## Detailed Design
+### Backup
+During backup, the main change is the backup-generated VSCs should not kept anymore.
+
+The reasons is we don't need them to ensure the snapshots clean up during backup deletion. Please reference to the [Backup Deletion section](#backup-deletion) section for detail.
+
+As a result, we can simplify the VS deletion logic in the backup. Before, we need to not only delete the VS, but also recreate a static VSC pointing a non-exiting VS.
+
+The deletion code in VS BackupItemAction can be simplify to the following:
+
+``` go
+	if backup.Status.Phase == velerov1api.BackupPhaseFinalizing ||
+		backup.Status.Phase == velerov1api.BackupPhaseFinalizingPartiallyFailed {
+		p.log.
+			WithField("Backup", fmt.Sprintf("%s/%s", backup.Namespace, backup.Name)).
+			WithField("BackupPhase", backup.Status.Phase).Debugf("Cleaning VolumeSnapshots.")
+
+		if vsc == nil {
+			vsc = &snapshotv1api.VolumeSnapshotContent{}
+		}
+
+		csi.DeleteReadyVolumeSnapshot(*vs, *vsc, p.crClient, p.log)
+		return item, nil, "", nil, nil
+	}
+
+
+func DeleteReadyVolumeSnapshot(
+	vs snapshotv1api.VolumeSnapshot,
+	vsc snapshotv1api.VolumeSnapshotContent,
+	client crclient.Client,
+	logger logrus.FieldLogger,
+) {
+	logger.Infof("Deleting Volumesnapshot %s/%s", vs.Namespace, vs.Name)
+	if vs.Status == nil ||
+		vs.Status.BoundVolumeSnapshotContentName == nil ||
+		len(*vs.Status.BoundVolumeSnapshotContentName) <= 0 {
+		logger.Errorf("VolumeSnapshot %s/%s is not ready. This is not expected.",
+			vs.Namespace, vs.Name)
+		return
+	}
+
+	if vs.Status != nil && vs.Status.BoundVolumeSnapshotContentName != nil {
+		// Patch the DeletionPolicy of the VolumeSnapshotContent to set it to Retain.
+		// This ensures that the volume snapshot in the storage provider is kept.
+		if err := SetVolumeSnapshotContentDeletionPolicy(
+			vsc.Name,
+			client,
+			snapshotv1api.VolumeSnapshotContentRetain,
+		); err != nil {
+			logger.Warnf("Failed to patch DeletionPolicy of volume snapshot %s/%s",
+				vs.Namespace, vs.Name)
+			return
+		}
+
+		if err := client.Delete(context.TODO(), &vsc); err != nil {
+			logger.Warnf("Failed to delete the VSC %s: %s", vsc.Name, err.Error())
+		}
+	}
+	if err := client.Delete(context.TODO(), &vs); err != nil {
+		logger.Warnf("Failed to delete volumesnapshot %s/%s: %v", vs.Namespace, vs.Name, err)
+	} else {
+		logger.Infof("Deleted volumesnapshot with volumesnapshotContent %s/%s",
+			vs.Namespace, vs.Name)
+	}
+}
+```
+
+### Restore
+
+#### Restore the VolumeSnapshotContent
+The current behavior of VSC restoration is that the VSC from the backup is restore, and the restored VS also triggers creating a new VSC dynamically.
+
+Two VSCs created for the same VS in one restore seems not right.
+
+Skip restore the VSC from the backup is not a viable alternative, because VSC may reference to a [snapshot create secret](https://kubernetes-csi.github.io/docs/secrets-and-credentials-volume-snapshot-class.html?highlight=snapshotter-secret-name#createdelete-volumesnapshot-secret).
+
+If the `SkipRestore` is set true in the restore action's result, the secret returned in the additional items is ignored too.
+
+As a result, restore the VSC from the backup, and setup the VSC and the VS's relation is a better choice.
+
+Another consideration is the VSC name should not be the same as the backed-up VSC's, because the older version Velero's restore and backup keep the VSC after completion.
+
+There's high possibility that the restore will fail due to the VSC already exists in the cluster.
+
+Multiple restores of the same backup will also meet the same problem.
+
+The proposed solution is using the restore's UID and the VS's name to generate sha256 hash value as the new VSC name. Both the VS and VSC RestoreItemAction can access those UIDs, and it will avoid the conflicts issues.
+
+The restored VS name also shares the same generated name.
+
+The VS-referenced VSC name and the VSC's snapshot handle name are in their status.
+
+Velero restore process purges the restore resources' metadata and status before running the RestoreItemActions.
+
+As a result, we cannot read these information in the VS and VSC RestoreItemActions.
+
+Fortunately, RestoreItemAction input parameters includes the `ItemFromBackup`. The status is intact in `ItemFromBackup`.
+
+``` go
+func (p *volumeSnapshotRestoreItemAction) Execute(
+	input *velero.RestoreItemActionExecuteInput,
+) (*velero.RestoreItemActionExecuteOutput, error) {
+	p.log.Info("Starting VolumeSnapshotRestoreItemAction")
+
+	if boolptr.IsSetToFalse(input.Restore.Spec.RestorePVs) {
+		p.log.Infof("Restore %s/%s did not request for PVs to be restored.",
+			input.Restore.Namespace, input.Restore.Name)
+		return &velero.RestoreItemActionExecuteOutput{SkipRestore: true}, nil
+	}
+
+	var vs snapshotv1api.VolumeSnapshot
+	if err := runtime.DefaultUnstructuredConverter.FromUnstructured(
+		input.Item.UnstructuredContent(), &vs); err != nil {
+		return &velero.RestoreItemActionExecuteOutput{},
+			errors.Wrapf(err, "failed to convert input.Item from unstructured")
+	}
+
+	var vsFromBackup snapshotv1api.VolumeSnapshot
+	if err := runtime.DefaultUnstructuredConverter.FromUnstructured(
+		input.ItemFromBackup.UnstructuredContent(), &vsFromBackup); err != nil {
+		return &velero.RestoreItemActionExecuteOutput{},
+			errors.Wrapf(err, "failed to convert input.Item from unstructured")
+	}
+
+	// If cross-namespace restore is configured, change the namespace
+	// for VolumeSnapshot object to be restored
+	newNamespace, ok := input.Restore.Spec.NamespaceMapping[vs.GetNamespace()]
+	if !ok {
+		// Use original namespace
+		newNamespace = vs.Namespace
+	}
+
+	if csiutil.IsVolumeSnapshotExists(newNamespace, vs.Name, p.crClient) {
+		p.log.Debugf("VolumeSnapshot %s already exists in the cluster. Return without change.", vs.Namespace+"/"+vs.Name)
+		return &velero.RestoreItemActionExecuteOutput{UpdatedItem: input.Item}, nil
+	}
+
+	newVSCName := generateSha256FromRestoreAndVsUID(string(input.Restore.UID), string(vsFromBackup.UID))
+	// Reset Spec to convert the VolumeSnapshot from using
+	// the dynamic VolumeSnapshotContent to the static one.
+	resetVolumeSnapshotSpecForRestore(&vs, &newVSCName)
+
+	// Reset VolumeSnapshot annotation. By now, only change
+	// DeletionPolicy to Retain.
+	resetVolumeSnapshotAnnotation(&vs)
+
+	vsMap, err := runtime.DefaultUnstructuredConverter.ToUnstructured(&vs)
+	if err != nil {
+		p.log.Errorf("Fail to convert VS %s to unstructured", vs.Namespace+"/"+vs.Name)
+		return nil, errors.WithStack(err)
+	}
+
+	p.log.Infof(`Returning from VolumeSnapshotRestoreItemAction with 
+		no additionalItems`)
+
+	return &velero.RestoreItemActionExecuteOutput{
+		UpdatedItem:     &unstructured.Unstructured{Object: vsMap},
+		AdditionalItems: []velero.ResourceIdentifier{},
+	}, nil
+}
+
+// generateSha256FromRestoreAndVsUID Use the restore UID and the VS UID to generate the new VSC name.
+// By this way, VS and VSC RIA action can get the same VSC name.
+func generateSha256FromRestoreAndVsUID(restoreUID string, vsUID string) string {
+	sha256Bytes := sha256.Sum256([]byte(restoreUID + "/" + vsUID))
+	return "vsc-" + hex.EncodeToString(sha256Bytes[:])
+}
+```
+
+#### Restore the VolumeSnapshot
+``` go
+// Execute restores a VolumeSnapshotContent object without modification
+// returning the snapshot lister secret, if any, as additional items to restore.
+func (p *volumeSnapshotContentRestoreItemAction) Execute(
+	input *velero.RestoreItemActionExecuteInput,
+) (*velero.RestoreItemActionExecuteOutput, error) {
+	if boolptr.IsSetToFalse(input.Restore.Spec.RestorePVs) {
+		p.log.Infof("Restore did not request for PVs to be restored %s/%s",
+			input.Restore.Namespace, input.Restore.Name)
+		return &velero.RestoreItemActionExecuteOutput{SkipRestore: true}, nil
+	}
+
+	p.log.Info("Starting VolumeSnapshotContentRestoreItemAction")
+
+	var vsc snapshotv1api.VolumeSnapshotContent
+	if err := runtime.DefaultUnstructuredConverter.FromUnstructured(
+		input.Item.UnstructuredContent(), &vsc); err != nil {
+		return &velero.RestoreItemActionExecuteOutput{},
+			errors.Wrapf(err, "failed to convert input.Item from unstructured")
+	}
+
+	var vscFromBackup snapshotv1api.VolumeSnapshotContent
+	if err := runtime.DefaultUnstructuredConverter.FromUnstructured(
+		input.ItemFromBackup.UnstructuredContent(), &vscFromBackup); err != nil {
+		return &velero.RestoreItemActionExecuteOutput{},
+			errors.Errorf(err.Error(), "failed to convert input.ItemFromBackup from unstructured")
+	}
+
+	// If cross-namespace restore is configured, change the namespace
+	// for VolumeSnapshot object to be restored
+	newNamespace, ok := input.Restore.Spec.NamespaceMapping[vsc.Spec.VolumeSnapshotRef.Namespace]
+	if ok {
+		// Update the referenced VS namespace to the mapping one.
+		vsc.Spec.VolumeSnapshotRef.Namespace = newNamespace
+	}
+
+	// Reset VSC name to align with VS.
+	vsc.Name = generateSha256FromRestoreAndVsUID(string(input.Restore.UID), string(vscFromBackup.Spec.VolumeSnapshotRef.UID))
+
+	// Reset the ResourceVersion and UID of referenced VolumeSnapshot.
+	vsc.Spec.VolumeSnapshotRef.ResourceVersion = ""
+	vsc.Spec.VolumeSnapshotRef.UID = ""
+
+	// Set the DeletionPolicy to Retain to avoid VS deletion will not trigger snapshot deletion
+	vsc.Spec.DeletionPolicy = snapshotv1api.VolumeSnapshotContentRetain
+
+	if vscFromBackup.Status != nil && vscFromBackup.Status.SnapshotHandle != nil {
+		vsc.Spec.Source.VolumeHandle = nil
+		vsc.Spec.Source.SnapshotHandle = vscFromBackup.Status.SnapshotHandle
+	} else {
+		p.log.Errorf("fail to get snapshot handle from VSC %s status", vsc.Name)
+		return nil, errors.Errorf("fail to get snapshot handle from VSC %s status", vsc.Name)
+	}
+
+	additionalItems := []velero.ResourceIdentifier{}
+	if csi.IsVolumeSnapshotContentHasDeleteSecret(&vsc) {
+		additionalItems = append(additionalItems,
+			velero.ResourceIdentifier{
+				GroupResource: schema.GroupResource{Group: "", Resource: "secrets"},
+				Name:          vsc.Annotations[velerov1api.PrefixedSecretNameAnnotation],
+				Namespace:     vsc.Annotations[velerov1api.PrefixedSecretNamespaceAnnotation],
+			},
+		)
+	}
+
+	vscMap, err := runtime.DefaultUnstructuredConverter.ToUnstructured(&vsc)
+	if err != nil {
+		return nil, errors.WithStack(err)
+	}
+
+	p.log.Infof("Returning from VolumeSnapshotContentRestoreItemAction with %d additionalItems",
+		len(additionalItems))
+	return &velero.RestoreItemActionExecuteOutput{
+		UpdatedItem:     &unstructured.Unstructured{Object: vscMap},
+		AdditionalItems: additionalItems,
+	}, nil
+}
+```
+
+
+### Backup Sync
+csi-volumesnapshotclasses.json, csi-volumesnapshotcontents.json, and csi-volumesnapshots.json are CSI-related metadata files in the BSL for each backup.
+
+csi-volumesnapshotcontents.json and csi-volumesnapshots.json are not needed anymore, but csi-volumesnapshotclasses.json is still needed.
+
+One concrete scenario is that a backup is created in cluster-A, then the backup is synced to cluster-B, and the backup is deleted in the cluster-B. In this case, we don't have a chance to create the VS and VSC needed VolumeSnapshotClass.
+
+The VSC deletion workflow proposed by this design needs to create the VSC first. If the VSC's referenced VolumeSnapshotClass doesn't exist in cluster, the creation of VSC will fail.
+
+As a result, the VolumeSnapshotClass should still be synced in the backup sync process.
+
+### Backup Deletion
+Two factors are worthy for consideration for the backup deletion change:
+* Because the VSCs generated by the backup are not synced anymore, and the VSCs generated during the backup will not be kept too. The backup deletion needs to generate a VSC, then deletes it to make sure the snapshots in the storage provider are clean too.
+* The VSs generated by the backup are already deleted in the backup process, we don't need a DeleteItemAction for the VS anymore. As a result, the `velero.io/csi-volumesnapshot-delete` plugin is unneeded.
+
+For the VSC DeleteItemAction, we need to generate a VSC. Because we only care about the snapshot deletion, we don't need to create a VS associated with the VSC.
+
+Create a static VSC, then point it to a pseudo VS, and reference to the snapshot handle should be enough.
+
+To avoid the created VSC conflict with older version Velero B/R generated ones, the VSC name is set to `vsc-uuid`.
+
+The following is an example of the implementation.
+``` go
+	uuid, err := uuid.NewRandom()
+	if err != nil {
+		p.log.WithError(err).Errorf("Fail to generate the UUID to create VSC %s", snapCont.Name)
+		return errors.Wrapf(err, "Fail to generate the UUID to create VSC %s", snapCont.Name)
+	}
+	snapCont.Name = "vsc-" + uuid.String()
+
+	snapCont.Spec.DeletionPolicy = snapshotv1api.VolumeSnapshotContentDelete
+
+	snapCont.Spec.Source = snapshotv1api.VolumeSnapshotContentSource{
+		SnapshotHandle: snapCont.Status.SnapshotHandle,
+	}
+
+	snapCont.Spec.VolumeSnapshotRef = corev1api.ObjectReference{
+		APIVersion: snapshotv1api.SchemeGroupVersion.String(),
+		Kind:       "VolumeSnapshot",
+		Namespace:  "ns-" + string(snapCont.UID),
+		Name:       "name-" + string(snapCont.UID),
+	}
+
+	snapCont.ResourceVersion = ""
+
+	if err := p.crClient.Create(context.TODO(), &snapCont); err != nil {
+		return errors.Wrapf(err, "fail to create VolumeSnapshotContent %s", snapCont.Name)
+	}
+
+	// Read resource timeout from backup annotation, if not set, use default value.
+	timeout, err := time.ParseDuration(
+		input.Backup.Annotations[velerov1api.ResourceTimeoutAnnotation])
+	if err != nil {
+		p.log.Warnf("fail to parse resource timeout annotation %s: %s",
+			input.Backup.Annotations[velerov1api.ResourceTimeoutAnnotation], err.Error())
+		timeout = 10 * time.Minute
+	}
+	p.log.Debugf("resource timeout is set to %s", timeout.String())
+
+	interval := 5 * time.Second
+
+	// Wait until VSC created and ReadyToUse is true.
+	if err := wait.PollUntilContextTimeout(
+		context.Background(),
+		interval,
+		timeout,
+		true,
+		func(ctx context.Context) (bool, error) {
+			tmpVSC := new(snapshotv1api.VolumeSnapshotContent)
+			if err := p.crClient.Get(ctx, crclient.ObjectKeyFromObject(&snapCont), tmpVSC); err != nil {
+				return false, errors.Wrapf(
+					err, "failed to get VolumeSnapshotContent %s", snapCont.Name,
+				)
+			}
+
+			if tmpVSC.Status != nil && boolptr.IsSetToTrue(tmpVSC.Status.ReadyToUse) {
+				return true, nil
+			}
+
+			return false, nil
+		},
+	); err != nil {
+		return errors.Wrapf(err, "fail to wait VolumeSnapshotContent %s becomes ready.", snapCont.Name)
+	}
+```
+
+## Security Considerations
+Security is not relevant to this design.
+
+## Compatibility
+In this design, no new information is added in backup and restore. As a result, this design doesn't have any compatibility issue.
+
+## Open Issues
+Please notice the CSI snapshot backup and restore mechanism not supporting all file-store-based volume, e.g. Azure Files, EFS or vSphere CNS File Volume. Only block-based volumes are supported.
+Refer to [this comment](https://github.com/vmware-tanzu/velero/issues/3151#issuecomment-2623507686) for more details.

design/Implemented/concurrent-backup-processing.md

@@ -0,0 +1,257 @@
+# Concurrent Backup Processing
+
+This enhancement will enable Velero to process multiple backups at the same time. This is largely a usability enhancement rather than a performance enhancement, since the overall backup throughput may not be significantly improved over the current implementation, since we are already processing individual backup items in parallel. It is a significant usability improvement, though, as with the current design, a user who submits a small backup may have to wait significantly longer than expected if the backup is submitted immediately after a large backup.
+
+## Background
+
+With the current implementation, only one backup may be `InProgress` at a time. A second backup created will not start processing until the first backup moves on to `WaitingForPluginOperations` or `Finalizing`. This is a usability concern, especially in clusters when multiple users are initiating backups. With this enhancement, we intend to allow multiple backups to be processed concurrently. This will allow backups to start processing immediately, even if a large backup was just submitted by another user. This enhancement will build on top of the prior parallel item processing feature by creating a dedicatede ItemBlock worker pool for each running backup. The pool will be created at the beginning of the backup reconcile, and the input channel will be passed to the Kubernetes backupper just like it is in the current release.
+
+The primary challenge is to make sure that the same workload in multiple backups is not backed up concurrently. If that were to happen, we would risk data corruption, especially around the processing of pod hooks and volume backup. For this first release we will take a conservative, high-level approach to overlap detection. Two backups will not run concurrently if there is any overlap in included namespaces. For example, if a backup that includes `ns1` and `ns2` is running, then a second backup for `ns2` and `ns3` will not be started. If a backup which does not filter namespaces is running (either a whole cluster backup or a non-namespace-limited backup with a label selector) then no other backups will be started, since a backup across all namespaces overlaps with any other backup. Calculating item-level overlap for queued backups is problematic since we don't know which items are included in a backup until backup processing has begun. A future release may add ItemBlock overlap detection, where at the item block worker level, the same item will not be processed by two different workers at the same time. This works together with workload conflict detection to further detect conflicts in a more granular level for shared resources between backups. Eventually, with a more complete understanding of individual workloads (either via ItemBlocks or some higher level model), the namespace level overlap detection may be relaxed in future versions.
+
+## Goals
+- Process multiple backups concurrently
+- Detect namespace overlap to avoid conflicts
+- For queued backups (not yet runnable due to concurrency limits or overlap), indicate the queue position in status
+
+## Non Goals
+- Handling NFS PVs when more than one PV point to the same underlying NFS share
+- Handling VGDP cancellation for failed backups on restart
+- Mounting a PVC for scenarios in which /tmp is too small for the number of concurrent backups
+- Providing a mechanism to identify high priority backups which get preferential treatment in terms of ItemBlock worker availability
+- Item-level overlap detection (future feature)
+- Providing the ability to disable namespace-level overlap detection once Item-level overlap detection is in place (although this may be supported in a future version).
+
+## High-Level Design
+
+### Backup CRD changes
+
+Two new backup phases will be added: `Queued` and `ReadyToStart`. In the Backup workflow, new backups will be moved to the Queued phase when they are added to the backup queue. When a backup is removed from the queue because it is now able to run, it will be moved to the `ReadyToStart` phase, which will allow the backup controller to start processing it.
+
+In addition, a new Status field, `QueuePosition`, will be added to track the backup's current position in the queue.
+
+### New Controller: `backupQueueReconciler`
+
+A new reconciler will be added, `backupQueueReconciler` which will use the current `backupReconciler` logic for reconciling `New` backups but instead of running the backup, it will move the Backup to the `Queued` phase and set `QueuePosition`.
+
+In addition, this reconciler will periodically reconcile all queued backups (on some configurable time interval) and if there is a runnable backup, remove it from the queue, update `QueuePosition` for any queued backups behind it, and update its phase to `ReadyToStart`.
+
+Queued backups will be reconciled in order based on `QueuePosition`, so the first runnable backup found will be processed. A backup is runnable if both of the following conditions are true:
+1) The total number of backups either `InProgress` or `ReadyToStart` is less than the configured number of concurrent backups.
+2) The backup has no overlap with any backups currently `InProgress` or `ReadyToStart` or with any `Queued` backups with a higher (i.e. closer to 1) queue position than this backup.
+
+### Updates to Backup controller
+
+The current `backupReconciler` will change its reconciling rules. Instead of watching and reconciling New backups, it will reconcile `ReadyToStart` backups. In addition, it will be configured to run in parallel by setting `MaxConcurrentReconciles` based on the `concurrent-backups` server arg.
+
+The startup (and shutdown) of the ItemBlock worker pool will be moved from reconciler startup to the backup reconcile, which will give each running backup its own dedicated worker pool. The per-backup worker pool will will use the existing `--item-block-worker-count` installer/server arg. This means that the maximum number of ItemBlock workers for the entire Velero pod will be the ItemBlock worker count multiplied by concurrentBackups. For example, if concurrentBackups is 5, and itemBlockWorkerCount is 6, then there will be, at most, 30 worker threads active, 5 dedicated to each InProgress backup, but this maximum will only be achieved when the maximum number of backups are InProgress. This also means that each InProgress backup will have a dedicated ItemBlock input channel with the same fixed buffer size.
+
+## Detailed Design
+
+### New Install/Server configuration args
+
+A new install/server arg, `concurrent-backups` will be added. This will be an int-valued field specifying the number of backups which may be processed concurrently (with phase `InProgress`). If not specified, the default value of 1 will be used.
+
+### Consideration of backup overlap and concurrent backup processing
+
+The primary consideration for running additional backups concurrently is the configured `concurrent-backups` parameter. If the total number of `InProgress` and `ReadyToStart` backups is equal to `concurrent-backups` then any `Queued` backups will remain in the queue.
+
+The second consideration is backup overlap. In order to prevent interaction between running backups (particularly around volume backup and pod hooks), we cannot allow two overlapping backups to run at the same time. For now, we will define overlap broadly -- requiring that two concurrent backups don't include any of the same namespaces. A backup for `ns1` can run concurrently with a backup for `ns2`, but a backup for `[ns1,ns2]` cannot run concurrently with a backup for `ns1`. One consequence of this approach is that a backup which includes all namespaces (even if further filtered by resource or label) cannot run concurrently with *any other backup*.
+
+When determining which queued backup to run next, velero will look for the next queued backup which has no overlap with any InProgress backup or any Queued backup ahead of it. The reason we need to consider queued as well as running backups for overlap detection is as follows.
+
+Consider the following scenario. These are the current not-completed backups (ordered from oldest to newest)
+1. backup1, includedNamespaces: [ns1, ns2], phase: InProgress
+2. backup2, includedNamespaces: [ns2, ns3, ns5], phase: Queued, QueuePosition: 1
+3. backup3, includedNamespaces: [ns4, ns3], phase: Queued, QueuePosition: 2
+4. backup4, includedNamespaces: [ns5, ns6], phase: Queued, QueuePosition: 2
+5. backup5, includedNamespaces: [ns8, ns9], phase: Queued, QueuePosition: 3
+
+Assuming `concurrent-backups` is 2, on the next reconcile, Velero will be able to start a second backup if there is one with no overlap. `backup2` cannot run, since `ns2` overlaps between it and the running `backup1`. If we only considered running overlap (and not queued overlap), then `backup3` could run now. It conflicts with the queued `backup2` on `ns3` but it does not conflict with the running backup. However, if it runs now, then when `backup1` completes, then `backup2` still can't run (since it now overlaps with running `backup3`on `ns3`), so `backup4` starts instead. Now when `backup3` completes, `backup2` still can't run (since it now conflicts with `backup4` on `ns5`). This means that even though it was the second backup created, it's the fourth to run -- providing worse time to completion than without parallel backups. If a queued backup has a large number of namespaces (a full-cluster backup for example), it would never run as long as new single-namespace backups keep being added to the queue.
+
+To resolve this problem we consider both running backups as well as backups ahead in the queue when resolving overlap conflicts. In the above scenario, `backup2` can't run yet since it overlaps with the running backup on `ns2`. In addition, `backup3` and `backup4` also can't run yet since they overlap with queued `backup2`. Therefore, `backup5` will run now. Once `backup1` completes, `backup2` will be free to run.
+
+### Backup CRD changes
+
+New Backup phases:
+```go
+const (
+	// BackupPhaseQueued means the backup has been added to the
+	// queue by the BackupQueueReconciler.
+	BackupPhaseQueued BackupPhase = "Queued"
+
+	// BackupPhaseReadyToStart means the backup has been removed from the
+	// queue by the BackupQueueReconciler and is ready to start.
+	BackupPhaseReadyToStart BackupPhase = "ReadyToStart"
+)
+```
+
+In addition, a new Status field, `queuePosition`, will be added to track the backup's current position in the queue.
+```go
+	// QueuePosition is the position held by the backup in the queue.
+	// QueuePosition=1 means this backup is the next to be considered.
+	// Only relevant when Phase is "Queued"
+	// +optional
+	QueuePosition int `json:"queuePosition,omitempty"`
+```
+
+### New Controller: `backupQueueReconciler`
+
+A new reconciler will be added, `backupQueueReconciler` which will reconcile backups under these conditions:
+1) Watching Create/Update for backups in `New` (or empty) phase
+2) Watching for Backup phase transition from `InProgress` to something else to reconcile all `Queued` backups
+2) Watching for Backup phase transition from `New` (or empty) to `Queued` to reconcile all `Queued` backups
+2) Periodic reconcile of `Queued` backups to handle backups queued at server startup as well as to make sure we never have a situation where backups are queued indefinitely because of a race condition or was otherwise missed in the reconcile on prior backup completion.
+
+The reconciler will be set up as follows -- note that New backups are reconciled on Create/Update, while Queued backups are reconciled when an InProgress backup moves on to another state or when a new backup moves to the Queued state. We also reconcile Queued backups periodically to handle the case of a Velero pod restart with Queued backups, as well as to handle possible edge cases where a queued backup doesn't get moved out of the queue at the point of backup completion or an error occurs during a prior Queued backup reconcile.
+
+```go
+func (c *backupOperationsReconciler) SetupWithManager(mgr ctrl.Manager) error {
+        // only consider Queued backups, order by QueuePosition
+	gp := kube.NewGenericEventPredicate(func(object client.Object) bool {
+		backup := object.(*velerov1api.Backup)
+		return (backup.Status.Phase == velerov1api.BackupPhaseQueued)
+	})
+	s := kube.NewPeriodicalEnqueueSource(c.logger.WithField("controller", constant.ControllerBackupOperations), mgr.GetClient(), &velerov1api.BackupList{}, c.frequency, kube.PeriodicalEnqueueSourceOption{
+		Predicates: []predicate.Predicate{gp},
+		OrderFunc: queuePositionOrderFunc,
+	})
+
+	return ctrl.NewControllerManagedBy(mgr).
+		For(&velerov1api.Backup{}, builder.WithPredicates(predicate.Funcs{
+				UpdateFunc: func(ue event.UpdateEvent) bool {
+					backup := ue.ObjectNew.(*velerov1api.Backup)
+					return backup.Status.Phase == "" || backup.status.Phase == velerov1api.BackupPhaseNew
+				},
+				CreateFunc: func(event.CreateEvent) bool {
+					return backup.Status.Phase == "" || backup.status.Phase == velerov1api.BackupPhaseNew
+				},
+				DeleteFunc: func(de event.DeleteEvent) bool {
+					return false
+				},
+				GenericFunc: func(ge event.GenericEvent) bool {
+					return false
+				},
+		})).
+		Watch(
+			&source.Kind{Type: &velerov1api.Backup{}},
+			&handler.EnqueueRequestsFromMapFunc{
+				ToRequests: handler.ToRequestsFunc(func(a handler.MapObject) []reconcile.Request {
+					backupList := velerov1api.BackupList{}
+					if err := p.List(ctx, backupList); err != nil {
+						p.logger.WithError(err).Error("error listing backups")
+						return
+					}
+					requests = []reconcile.request{}
+					// filter backup list by Phase=queued
+					// sort backup list by queuePosition
+					return requests
+				}),
+		},
+		builder.WithPredicates(predicate.Funcs{
+				UpdateFunc: func(ue event.UpdateEvent) bool {
+					oldBackup := ue.ObjectOld.(*velerov1api.Backup)
+					newBackup := ue.ObjectNew.(*velerov1api.Backup)
+					return oldBackup.Status.Phase == velerov1api.BackupPhaseInProgress &&
+						newBackup.Status.Phase != velerov1api.BackupPhaseInProgress ||
+						oldBackup.Status.Phase != velerov1api.BackupPhaseQueued &&
+						newBackup.Status.Phase == velerov1api.BackupPhaseQueued
+				},
+				CreateFunc: func(event.CreateEvent) bool {
+					return false
+				},
+				DeleteFunc: func(de event.DeleteEvent) bool {
+					return false
+				},
+				GenericFunc: func(ge event.GenericEvent) bool {
+					return false
+				},
+		}).
+		WatchesRawSource(s).
+		Named(constant.ControllerBackupQueue).
+		Complete(c)
+}
+```
+
+New backups will be queued: Phase will be set to `Queued`, and `QueuePosition` will be set to a int value incremented from the highest current `QueuePosition` value among Queued backups.
+
+Queued backups will be removed from the queue if runnable:
+1) If the total number of backups either InProgress or ReadyToStart is greater than or equal to the concurrency limit, then exit without removing from the queue.
+2) If the current backup overlaps with any InProgress, ReadyToStart, or Queued backup with `QueuePosition < currentBackup.QueuePosition` then exit without removing from the queue.
+3) If we get here, the backup is runnable. To resolve a potential race condition where an InProgress backup completes between reconciling the backup with QueuePosition `n-1` and reconciling the current backup with QueuePosition `n`, we also check to see whether there are any runnable backups in the queue ahead of this one. The only time this will happen is if a backup completes immediately before reconcile starts which either frees up a concurrency slot or removes a namespace conflict. In this case, we don't want to run the current backup since the one ahead of this one in the queue (which was recently passed over before the InProgress backup completed) must run first. In this case, exit without removing from the queue.
+4) If we get here, remove the backup from the queue by setting Phase to `ReadyToStart` and `QueuePosition` to zero. Decrement the `QueuePosition` of any other Queued backups with a `QueuePosition` higher than the current backup's queue position prior to dequeuing. At this point, the backup reconciler will start the backup.
+
+`if len(inProgressBackups)+len(pendingStartBackups) >= concurrentBackups`
+
+```
+	switch original.Status.Phase {
+	case "", velerov1api.BackupPhaseNew:
+		// enqueue backup -- set phase=Queued, set queuePosition=maxCurrentQueuePosition+1
+	}
+	// We should only ever get these events when added in order by the periodical enqueue source
+	// so as long as the current backup has not conflicts ahead of it or running, we should be good to
+	// dequeue
+	case "", velerov1api.BackupPhaseQueued:
+		// list backups, filter on Queued, ReadyToStart, and InProgress
+		// if number of InProgress backups + number of ReadyToStart backups >= concurrency limit, exit
+		// generate list of all namespaces included in InProgress, ReadyToStart, and Queued backups with
+		// queuePosition < backup.Status.QueuePosition
+		// if overlap found, exit
+		// check backups ahead of this one in the queue for runnability. If any are runnable, exit
+		// dequeue backup: set Phase to ReadyToStart, QueuePosition to 0, and decrement QueuePosition
+		// for all QueuedBackups behind this one in the queue
+	}
+
+```
+
+The queue controller will run as a single reconciler thread, so we will not need to deal with concurrency issues when moving backups from New to Queued or from Queued to ReadyToStart, and all of the updates to QueuePosition will be from a single thread.
+
+### Updates to Backup controller
+
+The Reconcile logic will be updated to respond to ReadyToStart backups instead of New backups:
+
+```
+@@ -234,8 +234,8 @@ func (b *backupReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctr
+        // InProgress, we still need this check so we can return nil to indicate we've finished processing
+        // this key (even though it was a no-op).
+        switch original.Status.Phase {
+-       case "", velerov1api.BackupPhaseNew:
+-               // only process new backups
++       case velerov1api.BackupPhaseReadyToStart:
++               // only process ReadyToStart backups
+        default:
+                b.logger.WithFields(logrus.Fields{
+                        "backup": kubeutil.NamespaceAndName(original),
+```
+
+In addition, it will be configured to run in parallel by setting `MaxConcurrentReconciles` based on the `concurrent-backups` server arg.
+
+```
+@@ -149,6 +149,9 @@ func NewBackupReconciler(
+ func (b *backupReconciler) SetupWithManager(mgr ctrl.Manager) error {
+        return ctrl.NewControllerManagedBy(mgr).
+                For(&velerov1api.Backup{}).
++                WithOptions(controller.Options{
++                       MaxConcurrentReconciles: concurrentBackups,
++               }).
+                Named(constant.ControllerBackup).
+                Complete(b)
+ }
+```
+
+The controller-runtime core reconciler logic already prevents the same resource from being reconciled by two different reconciler threads, so we don't need to worry about concurrency issues at the controller level.
+
+The workerPool reference will be moved from the backupReconciler to the backupRequest, since this will now be backup-specific, and the initialization code for the worker pool will be moved from the reconciler init into the backup reconcile. This worker pool will be shut down upon exiting the Reconcile method.
+
+### Resilience to restart of velero pod
+
+The new backup phases (`Queued` and `ReadyToStart`) will be resilient to velero pod restarts. If the velero pod crashes or is restarted, only backups in the `InProgress` phase will be failed, so there is no change to current behavior. Queued backups will retain their queue position on restart, and ReadyToStart backups will move to InProgress when reconciled.
+
+### Observability
+
+#### Logging
+
+When a backup is dequeued, an info log message will also include the wait time, calculated as `now - creationTimestamp`. When a backup is passed over due to overlap, an info log message will indicate which namespaces were in conflict.
+
+#### Velero CLI
+
+The `velero backup describe` output will include the current queue position for queued backups.

design/Implemented/handle-backup-of-volumes-by-resources-filters.md

@@ -86,7 +86,7 @@ volumePolicies:
     # capacity condition matches the volumes whose capacity falls into the range
     capacity: "0,100Gi"
     csi:
-      driver: aws.ebs.csi.driver
+      driver: ebs.csi.aws.com
       fsType: ext4
     storageClass:
     - gp2
@@ -174,7 +174,7 @@ data:
   - conditions:
       capacity: "0,100Gi"
       csi:
-        driver: aws.ebs.csi.driver
+        driver: ebs.csi.aws.com
         fsType: ext4
       storageClass:
       - gp2

design/Implemented/include-exclude-in-resource-policy.md

@@ -0,0 +1,82 @@
+# Proposal to add include exclude policy to resource policy
+
+This enhancement will allow the user to set include and exclude filters for resources in a resource policy configmap, so that
+these filters are reusable and the user will not need to set them each time they create a backup.
+
+## Background
+As mentioned in issue [#8610](https://github.com/vmware-tanzu/velero/issues/8610).  When there's a long list of resources 
+to include or exclude in a backup, it can be cumbersome to set them each time a backup is created.  There's a requirement to
+set these filters in a separate data structure so that they can be reused in multiple backups.
+
+## High-Level Design
+We may extend the data structure of resource policy to add `includeExcludePolicy`, which include the include and exclude filters 
+in the BackupSpec.  When the user creates a backup which references the resource policy config `velero backup create --resource-policies-configmap <configmap-name>`,
+the filters in "includeExcludePolicy" will take effect to filter the resources when velero collects the resources to backup.
+
+## Detailed Design
+
+### Data Structure
+The map `includeExcludePolicy` contains four fields `includedClusterScopedResources`, `excludedClusterScopedResources`, 
+`includedNamespaceScopedResources`,`excludedNamespaceScopedResources`.  These filters work exactly as the filters defined BackupSpec with
+the same names.  An example of the policy looks like:
+```yaml
+#omitted other irrelevant fields like 'version', 'volumePolicies'
+includeExcludePolicy:
+  includedClusterScopedResources:
+    - "cr"
+    - "crd"
+    - "pv"
+  excludedClusterScopedResources:
+    - "volumegroupsnapshotclass"
+    - "ingressclass"
+  includedNamespaceScopedResources:
+    - "pod"
+    - "service"
+    - "deployment"
+    - "pvc"
+  excludedNamespaceScopedResources:
+    - "configmap"
+```
+These filters are in the form of scoped include/exclude filters, which by design will not work with the "old" resource filters.
+Therefore, when a Backup references a resource policy configmap which has `includeExcludePolicy`, and at the same time it has 
+the "old" resource filters, i.e. `includedResources`, `excludedResources`, `includeClusterResources` set in the BackupSpec, the
+Backup will fail with a validation error.
+
+### Priorities 
+A user may set the include/exclude filters in Backupspec and also in the resource policy configmap.  In this case, the filters 
+in both the Backupspec and the resource policy configmap will take effect.  When there's a conflict, the filters in the Backupspec 
+will take precedence.  For example, if resource X is in the list of `includedNamespaceScopedResources` filter in the Backupspec, but 
+it's also in the list of `excludedClusterScopedResources` in the resource policy configmap, then resource X will be included in the backup.
+In this way, users can set the filters in the resource policy configmap to cover most of their use cases, and then override them 
+in the Backupspec when needed.
+
+### Implementation
+In addition to the data structure change, we will need to implement the following changes:
+1. A new function `CombineWithPolicy` will be added to the struct `ScopeIncludesExcludes`, which will combine the include/exclude filters
+in the resource policy configmap with the include/exclude filters in the Backupspec:  
+```go
+func (ie *ScopeIncludesExcludes) CombineWithPolicy(policy resourcepolicies.IncludeExcludePolicy) {
+	mapFunc := scopeResourceMapFunc(ie.helper)
+	for _, item := range policy.ExcludedNamespaceScopedResources {
+		resolvedItem := mapFunc(item, true)
+		if resolvedItem == "" {
+			continue
+		}
+		if !ie.ShouldInclude(resolvedItem) && !ie.ShouldExclude(resolvedItem) {
+			// The existing includeExcludes in the struct has higher priority, therefore, we should only add the item to the filter
+			// when the struct does not include this item and this item is not yet in the excludes filter.
+			ie.namespaceScopedResourceFilter.excludes.Insert(resolvedItem)
+		}
+		
+	}
+.....
+```
+This function will be called in the `kubernetesBackupper.BackupWithResolvers` function, to make sure the combined `ScopeIncludesExcludes` 
+filter will be assigned to the `ResourceIncludesExcludes` filter of the Backup request.
+
+2. Extra validation code will be added to the function `prepareBackupRequest` of `BackupReconciler` to check if there are "old"
+Resource filters in the BackupSpec when the Backup references a resource policy configmap which has `includeExcludePolicy`.
+
+## Alternatives Considered
+We may put `includeExcludePolicy` in a separate configmap, but it will require adding extra field to BackupSpec to reference the configmap,
+which is not necessary.

design/Implemented/move-gh-org.md

@@ -65,7 +65,7 @@ This page contains a pre-migration checklist for ensuring a repo migration goes
 
 #### Updating Netlify
 
-The settings for Netflify should remain the same, except that it now needs to be installed in the new repo. The instructions on how to install Netlify on the new repo are here: https://www.netlify.com/docs/github-permissions/.
+The settings for Netlify should remain the same, except that it now needs to be installed in the new repo. The instructions on how to install Netlify on the new repo are here: https://www.netlify.com/docs/github-permissions/.
 
 #### Communication strategy
 

design/Implemented/multiple-arch-build-with-windows.md

@@ -0,0 +1,122 @@
+# Multi-arch Build and Windows Build Support
+
+## Background
+
+At present, Velero images could be built for linux-amd64 and linux-arm64. We need to support other platforms, i.e., windows-amd64.  
+At present, for linux image build, we leverage Buildkit's `--platform` option to create the image manifest list in one build call. However, it is a limited way and doesn't fully support all multi-arch scenarios. Specifically, since the build is done in one call with the same parameters, it is impossbile to build images with different configurations (e.g., Windows build requires a different Dockerfile).   
+At present, Velero by default build images locally, or no image or manifest is pushed to registry. However, docker doesn't support multi-arch build locally. We need to clarify the behavior of local build.    
+
+## Goals
+- Refactor the `make container` process to fully support multi-arch build
+- Add Windows build to the existing build process
+- Clarify the behavior of local build with multi-arch build capabilities
+- Don't change the pattern of the final image tag to be used by users
+
+## Non-Goals
+- There may be some workarounds to make the multi-arch image/manifest fully available locally. These workarounds will not be adopted, so local build always build single-arch images
+
+## Local Build
+
+For local build, two values of `--output` parameter for `docker buildx build` are supported:
+- `docker`: a docker format image is built, but the image is only built for the platform (`<os>/<arch>`) as same as the building env. E.g., when building from linux-amd64 env, a single manifest of linux-amd64 is created regardless how the input parameters are configured.  
+- `tar`: one or more images are built as tarballs according to the input platform (`<os>/<arch>`) parameters. Specifically, one tarball is generated for each platform. The build process is the same with the `Build Separate Manifests` of `Push Build` as detailed below. Merely, the `--output` parameter diffs, as `type=tar;dest=<tarball generated path>`. The tarball is generated to the `_output` folder and named with the platform info, e.g., `_output/velero-main-linux-amd64.tar`.  
+
+## Push Build
+
+For push build, the `--output` parameter for `docker buildx build` is always `registry`. And build will go according to the input parameters and create multi-arch manifest lists.    
+
+### Step 1: Build Separate Manifests
+
+Instead of specifying multiple platforms (`<os>/<arch>`) to `--platform` option, we add multiple `container-%` targets in Makefile and each target builds one platform representively.  
+
+The goal here is to build multiple manifests through the multiple targets. However, `docker buildx build` by default creates a manifest list even though there is only one element in `--platform`. Therefore, two flags `--provenance=false` and `--sbom=false` will be set additionally to force `docker buildx build` to create manifests.  
+
+Each manifest has a unique tag, the OS type and arch is added to the tag, in the pattern `$(REGISTRY)/$(BIN):$(VERSION)-$(OS)-$(ARCH)`. For example, `velero/velero:main-linux-amd64`.  
+
+All the created manifests will be pushed to registry so that the all-in-one manifest list could be created.  
+
+### Step 2: Create All-In-One Manifest List
+
+The next step is to create a manifest list to include all the created manifests. This could be done by `docker manifest create` command, the tags created and pushed at Step 1 are passed to this command.  
+A tag is also created for the manifest list, in the pattern `$(REGISTRY)/$(BIN):$(VERSION)`. For example, `velero/velero:main`.  
+
+### Step 3: Push All-In-One Manifest List
+
+The created manifest will be pushed to registry by command `docker manifest push`.  
+
+## Input Parameters
+
+Below are the input parameters that are configurable to meet different build purposes during Dev and release cycle:
+- BUILD_OUTPUT_TYPE: the type of output for the build, i.e., `docker`, `tar`, `registry`, while `docker` and `tar` is for local build; `registry` means push build. Default value is `docker`  
+- BUILD_OS: which types of OS should be built for. Multiple values are accepted, e.g., `linux,windows`. Default value is `linux`  
+- BUILD_ARCH: which types of architecture should be built for. Multiple values are accepted, e.g., `amd64,arm64`. Default value is `amd64`  
+- BUILDX_INSTANCE: an existing buildx instance to be used by the build. Default value is <empty> which indicates the build to create a new buildx instance  
+
+## Windows Build
+
+Windows container images vary from Windows OS versions, e.g., `ltsc2022` for Windows server 2022 and `1809` for Windows server 2019. Images for different OS versions should be built separately.  
+Therefore, separate build targets are added for each OS version, like `container-windows-%`.  
+For the same reason, a new input parameter is added, `BUILD_WINDOWS_VERSION`. The default value is `ltsc2022`. Windows server 2022 is the only base image we will deliver officially, Windows server 2019 is not supported. In future, we may need to support Windows server 2025 base image.  
+For local build to tar, the Windows OS version is also added to the name of the tarball, e.g., `_output/velero-main-windows-ltsc2022-amd64.tar`.  
+
+At present, Windows container image only supports `amd64` as the architecture, so `BUILD_ARCH` is ignored for Windows.  
+
+The Windows manifests need to be annotated with os type, arch, and os version. This will be done through `docker manifest annotate` command.  
+
+## Use Malti-arch Images
+
+In order to use the images, the manifest list's tag should be provided to `velero install` command or helm, the individual manifests are covered by the manifest list. During launch time, the container engine will load the right image to the container according to the platform of the running node.  
+
+## Build Samples
+
+**Local build to docker**
+```
+make container
+```
+The built image could be listed by `docker image ls`.  
+
+**Local build for linux-amd64 and windows-amd64 to tar**
+```
+BUILD_OUTPUT_TYPE=tar BUILD_OS=linux,windows make container
+```
+Under `_output` directory, below files are generated:  
+```
+velero-main-linux-amd64.tar
+velero-main-windows-ltsc2022-amd64.tar
+``` 
+
+**Local build for linux-amd64, linux-arm64 and windows-amd64 to tar**
+```
+BUILD_OUTPUT_TYPE=tar BUILD_OS=linux,windows BUILD_ARCH=amd64,arm64 make container
+```
+Under `_output` directory, below files are generated:  
+```
+velero-main-linux-amd64.tar
+velero-main-linux-arm64.tar
+velero-main-windows-ltsc2022-amd64.tar
+```
+
+**Push build for linux-amd64 and windows-amd64**  
+Prerequisite: login to registry, e.g., through `docker login`  
+```
+BUILD_OUTPUT_TYPE=registry REGISTRY=<registry> BUILD_OS=linux,windows make container
+```
+Nothing is available locally, in the registry 3 tags are available:
+```
+velero/velero:main
+velero/velero:main-windows-ltsc2022-amd64
+velero/velero:main-linux-amd64
+```
+
+**Push build for linux-amd64, linux-arm64 and windows-amd64**  
+Prerequisite: login to registry, e.g., through `docker login` 
+```
+BUILD_OUTPUT_TYPE=registry REGISTRY=<registry> BUILD_OS=linux,windows BUILD_ARCH=amd64,arm64 make container
+```
+Nothing is available locally, in the registry 4 tags are available:
+```
+velero/velero:main
+velero/velero:main-windows-ltsc2022-amd64
+velero/velero:main-linux-amd64
+velero/velero:main-linux-arm64
+```