Supercharged GitHub Actions monitoring with OpenTelemetry at Elastic
summary
Built a custom OpenTelemetry Collector receiver that ingests GitHub Actions logs, traces and metrics, centralizes them to Elastic and Jaeger, and correlates spans with logs for root‑cause. Handles 242,000 workflows in 15 days by batching, backpressure and memory tuning, and works around GitHub API rate limits. Detects high‑level smells (deprecated set_output, command not found), fires alerts, creates Jira/GitHub issues, and measures CI SLOs for main‑branch health so failures and long‑running steps become actionable.
AI-written summary · about
chapters (15)
- 0:00Introduction and background
- 1:14What is CI/CD observability
- 4:02Our journey with GitHub Actions
- 8:13Demo of traces and logs
- 11:04Debugging the OpenTelemetry collector
- 12:51Collector configuration and metrics
- 14:39High-level GitHub workflow dashboard
- 15:40Log analysis and deprecated functions
- 17:05Detailed workflow trace in Kibana
- 19:11Creating actionable alerts and rules
- 22:00Service Level Objectives for CI
- 23:49Using AI for CI performance insights
- 25:13Challenges and lessons learned
- 27:19Future improvements and references
- 28:42Q&A and closing
related talks
44mSupercharged GitHub Actions monitoring with OpenTelemetry at Elastic
30mGoing Plaid: Striving for Speed of Light in CI pipelines
36mJMeter + GitHub = Understandable performance of your application
36mGitHub Actions: Open Source Workflow Automation
43mRewriting an Old CI/CD on GitHub Actions for 280 Contributors?
15mCentralized CI Monitoring with GitHub Actions
01h 47mDeveloper Workflow Automation with GitHub Actions
46mQuality Beyond Testing: Jenkins to GitHub Actions Migration: Optimized Processes & QA