Implementing SRE practices: SLI/SLO deep dive
summary
Explains how to turn reliability into measurable ratios (SLIs) like successful HTTP calls/total and simple math (50 of 100 = 50%), and how to set SLOs with concrete numbers and windows (example: 90% availability as measured at the load balancer, often rolling 30-day). Introduces error budget as the slack you can spend, lists responses when budget is exceeded or exhausted (speed releases, slow releases, instrument, stop releases), and stresses trusted monitoring and where metrics are measured.
AI-written summary · about
chapters (9)
- 0:00Introduction to SRE
- 5:44SRE vs DevOps
- 7:06SLI/SLO basics
- 11:43Types of SLIs
- 15:58Calculating SLIs
- 19:55Defining SLOs
- 23:51Error budgets explained
- 29:25Handling error budgets
- 34:52Getting started with SRE
