Alerting on error budget burn rate

This episode explains error-budget burn rate as a way to detect user-impacting degradation while limiting transient alert noise. Multi-window, multi-burn-rate logic balances fast detection with confidence that budget consumption is sustained ([00:00:00]-[00:02:07]). The transcript identifies neither the presenter nor an affiliation.

Key Points Covered

• Burn rate connects alerts to an SLO: It measures error-budget consumption relative to the objective and evaluation period ([00:00:00]-[00:02:07]).
• Long and short windows serve different purposes: Their combination can detect quickly while allowing an alert to reset after mitigation ([00:02:07]).
• Investigate the alert with adjacent evidence: The example compares SLI degradation and budget consumption with an increase in HTTP 500 logs ([00:04:12]-[00:05:14]).
• The threshold demonstration is internally inconsistent: The calculation gives 5.6 for one six-hour condition, but the UI walkthrough says to enter 6 ([00:01:04]-[00:02:07], [00:03:09]-[00:04:12]).

The 2021 UI appears to configure individual windows even though the episode describes multi-window logic. Verify current capabilities, syntax, formulas, and authoritative SRE threshold guidance rather than copying the values.

Full video: https://www.youtube.com/watch?v=t1BGo-Il1AM(opens in a new tab)