Build and simulate CI/CD pipelines — define stages and jobs, watch them run, analyse duration and failure patterns.
No stages yet — add a stage and some jobs, or load a preset.
Define pipeline stages in the Build tab first.
No run data yet — run the pipeline first.
| # | Status | Duration | Failed Jobs | Timestamp |
|---|---|---|---|---|
| No runs yet. | ||||
Continuous Integration merges developer changes frequently and runs automated builds/tests. Continuous Delivery ensures the build is always deployable. Continuous Deployment auto-deploys every passing build to production.
A stage is a logical phase (build, test, deploy). Stages run sequentially; a failed stage blocks later ones. Jobs within a stage run in parallel, each in its own isolated environment (container, VM, or runner).
Build outputs (compiled binaries, docker images, test reports) passed between stages. Good pipelines produce immutable artifacts — the same binary tested in staging is deployed to production, never rebuilt.
Reuse expensive downloads (npm modules, pip packages, Maven deps) across runs. Cache keys are typically hashed from lock files (package-lock.json) so a changed dep busts the cache automatically.
Named deployment targets (dev, staging, prod). Production environments typically require manual approval gates and have protection rules restricting who can trigger deploys.
Options include: redeploy previous artifact, git revert + redeploy, blue/green swap, canary weight reset. The fastest rollback is a feature flag flip — no redeploy needed.
Split test suites across N runners to reduce wall-clock time. Most CI systems support strategy.matrix (GitHub Actions) or parallel: (GitLab) for fan-out. Critical path analysis shows which job limits overall speed.
Credentials injected at runtime as env vars — never hardcoded or committed. Stored in the CI platform's secret store or external vaults (HashiCorp Vault, AWS Secrets Manager). Secrets are masked in logs automatically.