Skip to content

BUILD-10724: Import GitHub cache to S3 (migration fallback) #45

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
julien-carsique-sonarsource merged 1 commit into from
Mar 24, 2026
Merged

BUILD-10724: Import GitHub cache to S3 (migration fallback) #45

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
178 changes: 178 additions & 0 deletions .github/workflows/check-cache-migration.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,178 @@
---
name: Check cache migration (GitHub → S3)

on:
workflow_dispatch:
inputs:
environment:
description: Cache environment to check
required: true
default: prod
type: choice
options:
- prod
- dev

jobs:
check-migration:
runs-on: ubuntu-latest
name: Compare GitHub cache vs S3
permissions:
id-token: write
contents: read
actions: write # required to list GitHub cache entries and set the opt-out repository variable

steps:
- uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0

- name: Setup S3 cache credentials
id: aws-auth
uses: SonarSource/gh-action_cache/credential-setup@v1
with:
environment: ${{ inputs.environment }}

- name: List GitHub cache entries
id: gh-caches
shell: bash
env:
GITHUB_TOKEN: ${{ github.token }}
GITHUB_REPOSITORY: ${{ github.repository }}
run: |
# Fetch all GitHub cache entries (paginated, up to 10 000)
PAGE=1
PER_PAGE=100
ALL_ENTRIES="[]"
while true; do
RESPONSE=$(curl -s -f \
-H "Authorization: token $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
"https://api.github.com/repos/${GITHUB_REPOSITORY}/actions/caches?per_page=${PER_PAGE}&page=${PAGE}")
ENTRIES=$(echo "$RESPONSE" | jq '.actions_caches')
COUNT=$(echo "$ENTRIES" | jq 'length')
ALL_ENTRIES=$(echo "$ALL_ENTRIES $ENTRIES" | jq -s 'add')
if [[ "$COUNT" -lt "$PER_PAGE" ]]; then
break
fi
PAGE=$((PAGE + 1))
done

# Filter: include only target branches, exclude unwanted key patterns
FILTERED=$(echo "$ALL_ENTRIES" | jq '[
.[] |
select(
(.ref | test("^refs/heads/(main|master|branch-.+|dogfood-on-.+|feature/long/.+)$")) and
(.key | test("^(build-number-|mise-)") | not)
) |
{ ref: .ref, key: .key }
]')

TOTAL=$(echo "$ALL_ENTRIES" | jq 'length')
INCLUDED=$(echo "$FILTERED" | jq 'length')
echo "Total GitHub cache entries: $TOTAL"
echo "Included for migration check: $INCLUDED"

# Write to file for next step
echo "$FILTERED" > /tmp/gh_caches.json
echo "included-count=$INCLUDED" >> "$GITHUB_OUTPUT"

- name: List S3 cache objects
id: s3-objects
shell: bash
env:
AWS_ACCESS_KEY_ID: ${{ steps.aws-auth.outputs.AWS_ACCESS_KEY_ID }}
AWS_SECRET_ACCESS_KEY: ${{ steps.aws-auth.outputs.AWS_SECRET_ACCESS_KEY }}
AWS_SESSION_TOKEN: ${{ steps.aws-auth.outputs.AWS_SESSION_TOKEN }}
AWS_DEFAULT_REGION: eu-central-1
S3_BUCKET: sonarsource-s3-cache-${{ inputs.environment }}-bucket
run: |
# List all S3 objects with pagination handled internally by the CLI
aws s3 ls "s3://$S3_BUCKET/" --recursive \
| awk '{print $4}' \
| sort > /tmp/s3_keys.txt

S3_COUNT=$(wc -l < /tmp/s3_keys.txt)
echo "Total S3 cache objects: $S3_COUNT"
echo "s3-count=$S3_COUNT" >> "$GITHUB_OUTPUT"

- name: Compare and report
id: compare
shell: bash
env:
GITHUB_TOKEN: ${{ github.token }}
GITHUB_REPOSITORY: ${{ github.repository }}
run: |
INCLUDED=$(jq 'length' /tmp/gh_caches.json)
MIGRATED=0
MISSING=()

while IFS= read -r ENTRY; do
REF=$(echo "$ENTRY" | jq -r '.ref')
KEY=$(echo "$ENTRY" | jq -r '.key')
# S3 keys use full ref for push events (refs/heads/main/{key}) and short branch name for PR events (feat/someBranch/{key}).
# Check both forms since we cannot know which was used at save time.
REF_SHORT="${REF#refs/heads/}"
S3_KEY_FULL="${REF}/${KEY}"
S3_KEY_SHORT="${REF_SHORT}/${KEY}"
if grep -qxF "$S3_KEY_FULL" /tmp/s3_keys.txt || grep -qxF "$S3_KEY_SHORT" /tmp/s3_keys.txt; then
MIGRATED=$((MIGRATED + 1))
else
MISSING+=("$S3_KEY_SHORT")
fi
done < <(jq -c '.[]' /tmp/gh_caches.json)

echo ""
echo "========================================="
echo " Migration status: $MIGRATED / $INCLUDED"
echo "========================================="

if [[ "${#MISSING[@]}" -gt 0 ]]; then
echo ""
echo "Missing in S3 (${#MISSING[@]} entries):"
printf ' %s\n' "${MISSING[@]}"
fi

if [[ "$MIGRATED" -eq "$INCLUDED" && "$INCLUDED" -gt 0 ]]; then
echo ""
echo "All included GitHub cache entries are present in S3."
echo "migration-complete=true" >> "$GITHUB_OUTPUT"
else
echo "migration-complete=false" >> "$GITHUB_OUTPUT"
fi

- name: Set CACHE_IMPORT_GITHUB=false (migration complete)
if: steps.compare.outputs.migration-complete == 'true'
shell: bash
env:
GITHUB_TOKEN: ${{ github.token }}
GITHUB_REPOSITORY: ${{ github.repository }}
run: |
VARIABLE_NAME="CACHE_IMPORT_GITHUB"

# Check if variable already exists
STATUS=$(curl -s -o /dev/null -w "%{http_code}" \
-H "Authorization: token $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
"https://api.github.com/repos/${GITHUB_REPOSITORY}/actions/variables/${VARIABLE_NAME}")

if [[ "$STATUS" == "200" ]]; then
# Update existing variable
curl -s -f -X PATCH \
-H "Authorization: token $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
"https://api.github.com/repos/${GITHUB_REPOSITORY}/actions/variables/${VARIABLE_NAME}" \
-d '{"name":"'"$VARIABLE_NAME"'","value":"false"}'
echo "Updated repository variable $VARIABLE_NAME=false"
else
# Create new variable
curl -s -f -X POST \
-H "Authorization: token $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
"https://api.github.com/repos/${GITHUB_REPOSITORY}/actions/variables" \
-d '{"name":"'"$VARIABLE_NAME"'","value":"false"}'
echo "Created repository variable $VARIABLE_NAME=false"
fi

echo ""
echo "Migration complete — CACHE_IMPORT_GITHUB set to false"
echo "Import fallback will be disabled on next workflow runs."
echo "To re-enable migration mode, delete or set CACHE_IMPORT_GITHUB=true in repository variables."
147 changes: 147 additions & 0 deletions .github/workflows/test-cache-migration-gh2s3.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,147 @@
---
name: Test cache migration (GitHub → S3)

on:
push:
branches: [ master ]
pull_request:
workflow_dispatch:

jobs:
# Prerequisite: provision GitHub cache entries used by the test scenarios
provision-github-cache:
runs-on: github-ubuntu-latest-s
name: "Provision GitHub cache entries"
permissions:
contents: read
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Save GitHub cache entry (for import scenario)
run: mkdir -p ~/.cache/test-migration && echo "github-content" > ~/.cache/test-migration/test-file.txt
- uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
with:
path: ~/.cache/test-migration
key: test-migration-gh
- name: Save GitHub cache entry (for S3-hit scenario, to confirm it is NOT restored)
run: echo "github-content" > ~/.cache/test-migration/test-file.txt
- uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
with:
path: ~/.cache/test-migration
key: test-migration-s3hit

# Prerequisite: provision an S3 cache entry for the S3-hit scenario
provision-s3-cache:
runs-on: github-ubuntu-latest-s
name: "Provision S3 cache entry"
permissions:
id-token: write
contents: read
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Create S3 cache content
run: mkdir -p ~/.cache/test-migration && echo "s3-content" > ~/.cache/test-migration/test-file.txt
- name: Save to S3 cache
uses: ./
with:
path: ~/.cache/test-migration
key: test-migration-s3hit
environment: dev
backend: s3

# Scenario 1: S3 backend with import mode active (default behavior after migration begins)
test-s3-import-enabled:
needs: provision-github-cache
runs-on: github-ubuntu-latest-s
name: "S3 + import enabled (migration fallback)"
permissions:
id-token: write
contents: read
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Cache with migration fallback
id: cache
uses: ./
with:
path: ~/.cache/test-migration
key: test-migration-gh
environment: dev
backend: s3
# import-github-cache defaults to true — no need to set it explicitly
- name: Verify import succeeded
run: |
[[ "${{ steps.cache.outputs.cache-hit }}" == "true" ]] || { echo "ERROR: cache-hit is not true"; exit 1; }
[[ "$(cat ~/.cache/test-migration/test-file.txt)" == "github-content" ]] || { echo "ERROR: unexpected content, not restored from GitHub"; exit 1; }
rm -rf ~/.cache/test-migration # prevent saving to S3 so other scenarios don't find it

# Scenario 2: S3 backend with import mode explicitly disabled
test-s3-import-disabled:
needs: provision-github-cache
runs-on: github-ubuntu-latest-s
name: "S3 + import disabled (migration complete)"
permissions:
id-token: write
contents: read
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Cache without migration fallback
id: cache
uses: ./
with:
path: ~/.cache/test-migration
key: test-migration-gh
environment: dev
backend: s3
import-github-cache: 'false'
- name: Verify no import from GitHub
run: |
[[ "${{ steps.cache.outputs.cache-hit }}" != "true" ]] || { echo "ERROR: cache-hit is true but import should be disabled"; exit 1; }
test ! -f ~/.cache/test-migration/test-file.txt || { echo "ERROR: cache content was restored but import should be disabled"; exit 1; }

# Scenario 3: S3 backend with import disabled via env var (repo variable pattern)
test-s3-import-disabled-via-env:
needs: provision-github-cache
runs-on: github-ubuntu-latest-s
name: "S3 + import disabled via env var"
permissions:
id-token: write
contents: read
env:
CACHE_IMPORT_GITHUB: 'false'
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Cache without migration fallback (via env)
id: cache
uses: ./
with:
path: ~/.cache/test-migration
key: test-migration-gh
environment: dev
backend: s3
- name: Verify no import from GitHub
run: |
[[ "${{ steps.cache.outputs.cache-hit }}" != "true" ]] || { echo "ERROR: cache-hit is true but import should be disabled"; exit 1; }
test ! -f ~/.cache/test-migration/test-file.txt || { echo "ERROR: cache content was restored but import should be disabled"; exit 1; }

# Scenario 4: S3 hit — GitHub import step must be skipped entirely
test-s3-hit-skips-github-import:
needs: [ provision-github-cache, provision-s3-cache ]
runs-on: github-ubuntu-latest-s
name: "S3 hit → no GitHub import"
permissions:
id-token: write
contents: read
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
- name: Cache (S3 hit expected, import enabled)
id: cache
uses: ./
with:
path: ~/.cache/test-migration
key: test-migration-s3hit
environment: dev
backend: s3
# import-github-cache defaults to true — S3 hit should prevent GitHub from being tried
- name: Verify S3 content restored, not GitHub
run: |
[[ "${{ steps.cache.outputs.cache-hit }}" == "true" ]] || { echo "ERROR: expected S3 hit"; exit 1; }
[[ "$(cat ~/.cache/test-migration/test-file.txt)" == "s3-content" ]] || { echo "ERROR: unexpected content — GitHub import may have overridden S3"; exit 1; }
Loading
Loading