Shit, our Site is Down! Let's build a culture of reliability
summary
Concrete definitions and measurable examples decide: SLI examples include mean failure rate over 60s measured every 10s and median LCP on initial load over 5 days filtered by 4G users; tooling examples include ngrxfx plus PerformanceObserver sending metrics to Datadog; surprising numbers include tradeoffs like three nines to four nines yielding $900 value and five nines allowing ~5.5 minutes downtime per year; built practices include error budgets, quarterly SLOs, and post‑mortem cultures and whether they actually shift work from features to reliability.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 0:55Site reliability as UX
- 3:06Organizations and SRE
- 4:08Five myths of site reliability
- 6:38Embracing reliability and risk
- 7:20Defining service availability
- 8:46Culture of reliability with SLIs
- 10:29Measuring SLIs in Angular
- 11:33Service level objectives
- 12:17Planning for failure with error budgets
- 13:43Post-mortem culture
