How to Destroy an SRE
summary
Argues that overloaded alerting, brittle tooling and poor processes actively destroy SREs rather than topics alone: migration to Helm packaging with explicit composability and versioning for multi‑environment deploys, CI gated by GitLab CI/CD with integration tests, a staffing guideline of 50% engineering / 25% on‑call time, and pruning or fixing flaky tests causing a large share of CI failures. Shows where these practices save time, and where noise, tool churn and hostile code reviews cause burnout and attrition.
AI-written summary · about
chapters (9)
- 0:00Introduction and Session Overview
- 0:16Understanding the With Boost Use Case
- 2:01Challenges in Multi-Environment Deployment
- 3:21Helm for Packaging and Deployment
- 3:56Composability and Versioning with Helm
- 7:09Release Process and Automation
- 8:54Testing and Integration with GitLab CI/CD
- 13:06Customer Deployment and Rollout
- 13:42Q&A Session
related talks
47mPersistent SRE Antipatterns: Pitfalls on the Road to Creating a Successful
25mUnleashing the potential of SRE for modern operations
28mHow to Do SRE When You Have No SRE
52mPersistent SRE Antipatterns: Pitfalls On the...
30mWhen Is SRE Right For You?
05mSRE & Team Health
25mSRE in the Third Age
42mDemystifying SRE and DevOps