10.2.2 · Skill in Reliability engineering
SLOs and error budgets
Agreed reliability targets and the room for failure they allow.
A service level objective sets a target, such as 99.9% of requests succeeding. The error budget is the remaining 0.1%. When the budget runs out, reliability work takes priority over features.
Resources
2- bookStart hereImplementing SLOs (The Site Reliability Workbook, chapter 2) ↗Google SRE · 35 min
How to choose indicators and targets that match what users experience, and use the error budget to decide.
- bookGo deeperAlerting on SLOs (The Site Reliability Workbook, chapter 5) ↗Google SRE · 20 min
Burn-rate alerts that page people only when users are actually at risk.
Tools
none yetTools for this topic are being researched.