Free to read ยท Three levels
CI/CD: Evidence, Candidates, and Safe Delivery
Design a delivery process that connects source, tests, artifacts, authority, adoption, and recovery through explicit verifiable contracts.
Start learningIn this course
Eighteen developed lessons take a synthetic Parcel application from integration feedback to compatibility-aware release and recovery. The course teaches platform-independent reasoning with official GitLab, GitHub, SLSA, Reproducible Builds, OpenTelemetry, Google SRE, PostgreSQL, and DORA references. Each lesson includes concrete examples, contextual mistakes, and worked exercises.
No CI workflow is created or enabled, no remote pipeline is triggered, and no environment or database is contacted. Complete Python examples operate only on synthetic in-memory data and are verified locally; diagrams, shell fragments, SQL, and platform policy sketches are illustrative and non-runnable. A local model establishes only the stated invariant, not the behavior of a real delivery platform.
What you will learn
- Define integration, delivery, and deployment promises precisely.
- Design pipeline data and scheduling dependencies around immutable revisions.
- Preserve test failures and promote identified artifacts through reviewed decisions.
- Separate cache acceleration, credentials, configuration, and environment authority.
- Assess compatibility, progressive exposure, consumer verification, and recovery.
- Use explicit event definitions and balanced delivery metrics for improvement.
Your learning path
Three levels
Integration and candidate foundations
Connect events, source, graph dependencies, tests, and retained packages.
- Integration, delivery, and deployment are different promises
- Dependencies, parallelism, and pipeline graphs
- Source events and the revision under test
- Reproducible inputs and deterministic packaging
- Tests, exit status, and useful release gates
- Artifacts, manifests, and release identity
Promotion and environment contracts
Separate acceleration, authority, configuration, compatibility, and observed behavior.
- Dependency caches without hidden correctness
- Secrets and trust boundaries in delivery
- Environments, configuration, and candidate promotion
- Progressive delivery and bounded exposure
- Database changes and compatibility windows
- Release observability and feedback
Integrity, coordination, and improvement
Evaluate supply-chain claims, runtime uncertainty, release governance, and recovery.
- Supply-chain verification at the consumer boundary
- Runner isolation and pipeline reliability
- Concurrency, stale releases, and idempotent retries
- Policy, approvals, and controlled exceptions
- Rollback, roll-forward, and recovery rehearsal
- Delivery measurement and improvement experiments
Who this is for
- Developers learning to design reliable automated feedback and release paths.
- Engineers reviewing artifact promotion, release authority, compatibility, and operational evidence.
Before you start
- Basic source-control, command-line, testing, and application configuration knowledge.
- Ability to read small Python programs and JSON; no CI service account or deployment access is required.