Restructures CI/CD pipeline to eliminate redundant Docker image builds
across parallel test workflows. Previously, every PR triggered 5 separate
builds of identical images, consuming compute resources unnecessarily and
contributing to registry storage bloat.
Registry storage was growing at 20GB/week due to unmanaged transient tags
from multiple parallel builds. While automated cleanup exists, preventing
the creation of redundant images is more efficient than cleaning them up.
Changes CI/CD orchestration so docker-build.yml is the single source of
truth for all Docker images. Integration tests (CrowdSec, Cerberus, WAF,
Rate Limiting) and E2E tests now wait for the build to complete via
workflow_run triggers, then pull the pre-built image from GHCR.
PR and feature branch images receive immutable tags that include commit
SHA (pr-123-abc1234, feature-dns-provider-def5678) to prevent race
conditions when branches are updated during test execution. Tag
sanitization handles special characters, slashes, and name length limits
to ensure Docker compatibility.
Adds retry logic for registry operations to handle transient GHCR
failures, with dual-source fallback to artifact downloads when registry
pulls fail. Preserves all existing functionality and backward
compatibility while reducing parallel build count from 5× to 1×.
Security scanning now covers all PR images (previously skipped),
blocking merges on CRITICAL/HIGH vulnerabilities. Concurrency groups
prevent stale test runs from consuming resources when PRs are updated
mid-execution.
Expected impact: 80% reduction in compute resources, 4× faster
total CI time (120min → 30min), prevention of uncontrolled registry
storage growth, and 100% consistency guarantee (all tests validate
the exact same image that would be deployed).
Closes #[issue-number-if-exists]
The "Save Docker Image as Artifact" and "Upload Image Artifact" steps
were running even when skip_build=true, causing CI failures on Renovate
dependency update PRs.
Add skip_build check to artifact saving step condition
Add skip_build check to artifact upload step condition
Aligns artifact steps with existing build skip logic
- Updated Docker Compose files to use digest-pinned images for CI contexts.
- Enhanced Dockerfile to pin Go tool installations and verify external downloads with SHA256 checksums.
- Added Renovate configuration for tracking Go tool versions and digest updates.
- Introduced a new design document outlining the architecture and data flow for dependency tracking.
- Created tasks and requirements documentation to ensure compliance with the new digest pinning policy.
- Updated security documentation to reflect the new digest pinning policy and exceptions.
Add missing emergency token environment variable to all E2E test workflows to
fix security teardown failures in CI. Without this token, the emergency reset
endpoint returns 501 "not configured", causing test teardown to fail and
leaving ACL enabled, which blocks 83 subsequent tests.
Changes:
Add CHARON_EMERGENCY_TOKEN to docker-build.yml test-image job
Add CHARON_EMERGENCY_TOKEN to e2e-tests.yml e2e-tests job
Add CHARON_EMERGENCY_TOKEN to playwright.yml playwright job
Verified:
Docker build strategy already optimal (build once, push to both GHCR + Docker Hub)
Testing strategy correct (test once by digest, validates both registries)
All workflows now have environment parity with local development setup
Requires GitHub repository secret:
Name: CHARON_EMERGENCY_TOKEN
Value: 64-char hex token (e.g., from openssl rand -hex 32)
Related:
Emergency endpoint rate limiting removal (proper fix)
Local emergency token configuration (.env, docker-compose.local.yml)
Security test suite teardown mechanism
Refs #550
GitHub Actions doesn't allow secrets context in step if expressions.
Add HAS_DOCKERHUB_TOKEN env var at job level that evaluates the secret
existence, then reference that env var in step conditions.
Fixes: "Unrecognized named-value: 'secrets'" workflow validation error
Publish Docker images to both Docker Hub (docker.io/wikid82/charon) and
GitHub Container Registry (ghcr.io/wikid82/charon) for maximum reach.
Add Docker Hub login with secret existence check for graceful fallback
Update docker/metadata-action to generate tags for both registries
Add Cosign keyless signing for both GHCR and Docker Hub images
Attach SBOM to Docker Hub via cosign attach sbom
Add Docker Hub signature verification to supply-chain-verify workflow
Update README with Docker Hub badges and dual registry examples
Update getting-started.md with both registry options
Supply chain security maintained: identical tags, signatures, and SBOMs
on both registries. PR images remain GHCR-only.
On PRs, images are loaded locally but not pushed to registry.
Add --pull=never to prevent Docker from trying to fetch the
image from ghcr.io, which fails with 'manifest unknown'.
Modified 4 docker commands:
- Caddy version check (docker run)
- Caddy binary extraction (docker create)
- CrowdSec version check (docker run)
- CrowdSec binary extraction (docker create)
Debian-based image takes longer to start than Alpine due to:
- Larger base image
- gosu and CrowdSec built from source
- Additional package dependencies
Increase timeout from 120s to 180s to accommodate slower startup.
- Updated WafConfig.tsx to correct regex for common bad bots.
- Modified cerberus_integration.sh to use curl instead of wget for backend readiness check.
- Changed coraza_integration.sh to utilize curl for checking httpbin backend status.
- Updated crowdsec_startup_test.sh to use curl for LAPI health check.
- Replaced wget with curl in install-go-1.25.5.sh for downloading Go.
- Modified rate_limit_integration.sh to use curl for backend readiness check.
- Updated waf_integration.sh to replace wget with curl for checking httpbin backend status.
Migrated all Docker stages from Alpine 3.23 to Debian Trixie (13) to
address critical CVE in Alpine's gosu package and improve security
update frequency.
Key changes:
Updated CADDY_IMAGE to debian:trixie-slim
Added gosu-builder stage to compile gosu 1.17 from source with Go 1.25.6
Migrated all builder stages to golang:1.25-trixie
Updated package manager from apk to apt-get
Updated user/group creation to use groupadd/useradd
Changed nologin path from /sbin/nologin to /usr/sbin/nologin
Security impact:
Resolved gosu Critical CVE (built from source eliminates vulnerable Go stdlib)
Reduced overall CVE count from 6 (bookworm) to 2 (trixie)
Remaining 2 CVEs are glibc-related with no upstream fix available
All Go binaries verified vulnerability-free by Trivy and govulncheck
Verification:
E2E tests: 243 passed (5 pre-existing failures unrelated to migration)
Backend coverage: 87.2%
Frontend coverage: 85.89%
Pre-commit hooks: 13/13 passed
TypeScript: 0 errors
Refs: CVE-2026-0861 (glibc, no upstream fix - accepted risk)
Feature branch pushes were failing to save artifacts because the image
was pushed to GHCR but not loaded locally. Multi-platform builds
cannot use load:true, so feature branch pushes now build single-platform.
Update branch triggers and downstream workflow logic to support all
branches defined in docker-build.yml (main, development, feature/**).
Changes:
docker-build.yml: Expand branch glob to feature/**, use branch-based tags
playwright.yml: Replace is_beta_push with generic is_push detection
security-pr.yml: Same branch-agnostic pattern
supply-chain-pr.yml: Same pattern, skip PR comments for push events
The workflows now support any push that triggers docker-build:
main branch → tag: latest
development branch → tag: dev
feature/* branches → tag: {branch-name}
Pull requests → tag: pr-{number}
Dynamic artifact naming:
Push events: push-image (shared across all branches)
Pull requests: pr-image-{number}
This ensures CI/CD pipelines work for stable releases, bug fixes,
and new feature development without hardcoded branch names.
Separate PR-specific tests from docker-build.yml into dedicated workflows
that trigger via workflow_run. This creates a cleaner CI architecture where:
playwright.yml: E2E tests triggered after docker-build completes
security-pr.yml: Trivy binary scanning for PRs
supply-chain-pr.yml: SBOM generation + Grype vulnerability scanning
The skip condition used github.ref to detect the beta-release branch,
but for PRs github.ref is "refs/pull/N/merge", not the branch name.
Added github.head_ref to env variables for PR branch detection
Updated condition to check both REF and HEAD_REF
This ensures E2E tests run for PRs from feature/beta-release branch
The DNS provider API endpoints were returning 404 in CI because the
encryption service failed to initialize with an invalid key.
Changed CHARON_ENCRYPTION_KEY from plain text to valid base64 string
Key "dGVzdC1lbmNyeXB0aW9uLWtleS1mb3ItY2ktMzJieXQ=" decodes to 32 bytes
Without valid encryption key, DNS provider routes don't register
This was causing all dns-provider-types.spec.ts tests to fail
Root cause: AES-256-GCM requires exactly 32 bytes for the key
