Artifact Management in Pipelines
You'll learn how CI/CD pipelines store, version, and promote build outputs across deployment stages.
What Are Build Artifacts?
When your pipeline builds code, it produces artifacts—compiled binaries, Docker images, JAR files, ZIP archives, or any packaged output ready for deployment. Think of artifacts like finished products coming off an assembly line: they need to be labeled, stored safely, and moved to the right destination.
The Three Pillars of Artifact Management
1. Storage
Artifacts live in artifact repositories (Nexus, Artifactory, cloud storage buckets). After the build stage succeeds, the pipeline uploads artifacts to these central locations. This keeps your pipeline fast—subsequent stages download pre-built artifacts instead of rebuilding from scratch.
2. Versioning
Every artifact gets a unique identifier—often a version number, commit SHA, or build ID. Version 1.2.3-abc123 tells you exactly which code produced that artifact. This traceability is crucial: if production breaks, you know precisely which artifact to investigate or rollback to.
3. Promotion
Artifacts "graduate" through environments: dev → staging → production. The same artifact (not a rebuild) moves forward. Your pipeline might tag artifact app:1.2.3 as app:staging-approved after passing tests, then app:production-ready after manual approval. You're promoting the tested artifact, not creating a new one.
The Pattern in Practice
A typical flow: commit triggers build → artifact uploaded with version tag → test stage pulls that artifact → if tests pass, tag artifact as "staging-ready" → deployment stage uses that specific version. No code changes between environments means predictable deployments.
Key Takeaway: Artifact management ensures build outputs are stored centrally with clear versions and promoted unchanged through pipeline stages, guaranteeing consistency from testing to production.