Stopping Pipeline Drift with GitHub Reusable Workflows
Stop duplicating CI/CD YAML across your repositories. Learn how to use GitHub Reusable Workflows to centralize pipeline logic, manage versioning, and eliminate pipeline drift.
ReadMeFeed / Community knowledge
Real questions. Useful conversations. Find the people who know your stack.
Stop duplicating CI/CD YAML across your repositories. Learn how to use GitHub Reusable Workflows to centralize pipeline logic, manage versioning, and eliminate pipeline drift.
A concise architecture note for deploying a minimal, secure self‑hosted GitHub Actions runner, covering requirements, design, trust boundaries, ops checks, failure modes, and when to redesign.
Learn how to diagnose and fix GitHub Actions jobs that stay in the 'Queued' state, focusing on runner label mismatches, concurrency bottlenecks, and offline self-hosted agents.
Stop duplicating your GitHub Actions jobs. Learn how to use Matrix Strategies to test across multiple OS and runtime versions efficiently using include, exclude, and fail-fast settings.
Learn how to cut CI build times by 30–70% with GitHub Actions caching. This guide covers key concepts, a concrete example, trade‑offs, and actionable steps to implement and verify a cache‑enabled workflow.
Learn when to use 'push' vs 'pull_request' triggers in GitHub Actions to avoid redundant runs and ensure your main branch remains stable through merge-commit validation.
actions/cache speeds up GitHub Actions by restoring your package manager's download store between runs — but only if your key includes a lockfile hash.
Stop running sequential tests. Learn how to use GitHub Actions Matrix Strategies to parallelize your test suites across multiple OS and language versions for faster feedback.
Goal: achieve identical schema migration outcomes when running Supabase CLI locally via supabase start and in a CI pipeline that uses supabase db push against a temporary project. Constraints: the CLI does not automatically remove ephemeral Supabase projects created for CI, requiring manual cleanup to avoid quota limits, and the Docker images pulled by supab