Skip to main content

Pedram Sarani

SRE / DevOps Notes

FA
Back to learning paths

Learning path

GitLab Platform Delivery at Scale

A practical path for designing GitLab CI/CD pipelines, runner capacity, cache strategy, release gates, and troubleshooting workflows for shared delivery platforms.

Outcomes

  • Design pipeline topology that separates fast merge request feedback from slower governance and release checks.
  • Operate GitLab runners with capacity, isolation, cache correctness, and timeout discipline.
  • Build release gates and rollback checks without turning CI/CD into an opaque approval machine.

Milestones

1) Pipeline architecture and feedback loops

  • DAG design with explicit job dependencies
  • Rules, path filters, and fast-fail checks for merge requests
  • Artifact and cache boundaries that avoid stale or cross-branch state

2) Runner operations and capacity

  • Executor choice, runner isolation, and privileged job controls
  • Queue time, concurrency limits, cache hit rate, and retry causes
  • Dashboard and alert patterns for saturated shared runners

3) Delivery governance and incident readiness

  • Protected environments, approvals, and release evidence
  • Helm and Argo CD handoff points from pipeline to cluster
  • Rollback workflow, deployment freeze behavior, and incident communication