SRE Theory vs. Practice: A Song of Ice and TireFire
summary
A postmortem of polished conference rhetoric vs messy reality: a touted “perfect” SRE setup actually hides chaos engineering used as blame insurance, dashboards that prove any correlation, and a Bay‑Area hiring fetish. Offers concrete failure stories — MySQL used as a distributed lock that collapses at high QPS, a playbook for a 10x load plan that still misses organizational constraints, and a discovered unit cost — cost per install: $0.04 — that flips project priorities when measured.
AI-written summary · about
chapters (31)
- 0:00Introduction
- 0:59Corey Quinn
- 1:25What is SRE
- 3:02Blameless Postmortems
- 3:54Chaos Engineering
- 4:50Dashboards
- 5:27Our greatest asset
- 6:38Rabbit Punch
- 6:54Socks
- 7:21Vendors
- 8:07Core Competency
- 8:52Bay Area
- 9:47Performance System
- 11:42Dont Be John
- 12:51Pacific Islands
- 13:39First Principles
- 13:58Cargo Cult
- 14:30Finding Context
- 15:34Industry Zeitgeist
- 15:56Blockchain
- 16:13SRE
- 16:36Martin Fowler
- 16:50The value of a maturity model
- 17:05A simple graph
- 17:46Tracking progress
- 18:47Leaving Google
- 19:34Collecting metrics
- 20:51Warning
- 26:52Data maturity
- 27:31Ask for help
- 28:13Netflix culture deck
related talks
18mThe Good, the Bad and the Ugly: The 3 Learnings of an SRE
25mMental Models for SREs
29mThe Curious SRE: Cultivating a Mindset for System Resilience
13mSucceeding as the Lone SRE in a Small Team
38mSRE Saga: The Song of Heroes and Villains
15mThe Biases Confronting SREs
48mThe Secret Lives of SREs - Controlling the Costs of Coordination across Remote
40m100 Teams, 100 Ways to Fail