Story of Seeking Sustainable SRE
summary
SoundCloud has ~100 million users but only 100 engineers, so microservices outnumber people and Google-style SRE fails when copied directly. The company builds automation, runs Kubernetes and Prometheus for monitoring, removes a central SRE by making developers run their own services and be on-call, and uses a site-wide availability target and soft error budgets on dashboards. The result is fewer pages, more automation, and a workable trade-off between shipping and reliability.
AI-written summary · about
chapters (6)
- 0:00Start
- 13:35Management support. • Nuclear option: Return the pager.
- 18:08Minimal size of a dedicated SRE team: 8. • Feasible percentage of all engineers in SRE: 5%? 10%?
- 18:59Less than 50% ops work.
- 21:43SRE is what happens when you ask a software engineer to design an operations team.
- 24:40Error budgets?
related talks
25mSRE in the Third Age
41mAn Introduction to SRE at Google (with DevOps Focus)
32mHow Google SRE and Developers Work Together
40mHow Google SRE and developers work together
27mHow Google SRE and developers work together
27mUnified Theory of SRE
05mHow do Devops and SRE relate?
13mSRE: Lessons from a Parallel Universe