Release Engineering as a Service - The Art of Integrating with K8s
summary
Pivotal’s release engineering team replaces brittle bash scripts with Ruby and Go, splits and templatizes large YAML pipelines, and automates environments with Concourse pipelines and Terraform for on‑demand environments. A flakiness study of 112 runs -> 55% success identifies the worst jobs to fix; parallelizing jobs cuts a 2h task to 15m and reduces full pipeline time from 12h to 4h, enabling builds every ~10 days and 48‑hour bug patch turnarounds.
AI-written summary · about
chapters (18)
- 0:00What is release engineering in our context
- 1:06What is Concourse?
- 1:48Terminologies
- 3:42No Testing
- 6:58Why is duplicate yaml
- 8:14DRY YAML: Decouple
- 8:38ERB Embedded
- 9:12Cons: ERB (Embedded Ruby)
- 10:07Cons: anchors & aliases
- 10:44Attempt 3
- 11:10templates +
- 11:54Environment contention
- 14:35Pipelines Take
- 15:01Pipelines Are Slow
- 16:32Pipelines Are Flaky
- 17:50No Integration Process
- 19:02Releng as a Service!
- 20:42Open Source Licensing
related talks
31mRelease Engineering as a Service - The Art of Integrating with K8s
17mA Brief History of Release Engineering
54mFrom Code to Customer
38mSIG Release - Deep Dive: Release Engineering Subproject
33mA Day in the Life of Kubernetes Release with Tools, Challenges, and Operations
19mKeynote: Releasing a More Mature and Stable Kubernetes
20mRelease Engineering For A Polyglot Team
20mStreamlining Release Management for DevOps Success