A Guide to Event-driven SRE-inspired DevOps
summary
Claims that 95% pipeline maintenance wastes engineering time and that manual validation consumes ~80% of wait time, then shows Keptn as an event-driven control plane on Kubernetes that separates process (shipyard) from tools (uniform). Uses Keptn CLI commands like keptn send event new-artifact, integrates with GitLab, Helm, JMeter, Dynatrace/Prometheus, runs on EKS, and enforces quality gate 90% threshold (demo build scored 86 and triggered rollback). Shows automated remediation workflows that recheck SLI/SLOs and escalate if fixes fail.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:20Multistage delivery
- 4:45Monolithic pipelines
- 6:13Eventdriven delivery
- 10:24Demo
- 13:33Manual Deployment Validation
- 14:42Quality Gate Approach
- 19:29Remediation Approach
- 22:34SRE vs DevOps
- 24:15First steps to DevOps
- 26:16Most voted question
- 27:07Managing debugging issues
- 28:38Can Captain replace Jenkins
- 29:52How does Captain integrate with other tools
- 31:31Difference between containers
- 31:56Is it possible to use Captain only to analyze data from Prometheus
- 32:46Wrapup
related talks
36mKeptn Beyond 1.0: Sailing into the Future
29mObservability Driven Automation: Beyond GitOps with Keptn
19mWorking Your Way Through GitOps with Keptn
32mEvent-based Continuous Delivery: The End of the Monolithic Pipeline
18mPerformance Testing in Keptn Using K6
27mContinuous Reliability. How?
47mModern continous delivery on Azure Kubernetes Service
29mThanks Keptn Obvious! Observable SLOs with Prometheus and Keptn