Applying SRE Principles to CI/CD
summary
Buildkite’s CI/CD is analyzed through SRE practices with concrete measures: an observed 9413 days retrying failed steps across users in one month, agent-backed pipelines that scale to thousands of agents, and instrumentation via OpenTelemetry sent to Datadog/Prometheus/Honeycomb. Practical SLO examples include builds starting within 30s, commits tested in 5 minutes, and a test-suite reliability >87% SLO. Test Analytics identifies flaky and slow tests (one fell 35s→3s), error budgets guide when to pause features and fix pipelines.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:55CI/CD complexity
- 3:22Flaky tests and failed builds
- 10:00SRE for reliable CI/CD
- 14:57Measuring CI/CD reliability
- 17:45SLOs, SLIs, error budgets
- 20:11Build performance optimization
- 21:58Test suite health tracking
- 24:07Error budget management
- 25:32SRE-driven CI/CD benefits
- 26:14Q&A
related talks
35mApplying SRE Principles to CI/CD
36mThinking like a site reliability engineer to improve CI
27mContinuous Reliability. How?
32mPractical Tips & Tricks for CI/CD Success
59mBridging the Gap Between Ops and Developers with CI/CD
26mIncreasing the collaboration between SREs and Developers
27mGetting Started with SRE
21mSRE, what is it all about?!