Multi-window burn-rate alerting
On This Lesson 4 topics
This lesson is planned. The outline below lists the topics it will cover. The full lesson has not been published yet.
In this lesson
page quickly on real outages and not on blips.
Planned topics
- Burn rate: how fast the budget is consumed relative to an even spend. Burn rate 1 uses exactly the budget over the window.
- Alert on burn rate, not raw error rate, using a long window for significance and a short window to confirm it is still happening.
- Common pairing from the SRE literature: fast burn (for example 14.4× over 1 h with a 5 min check) pages; slow burn (for example 1× over 3 days) opens a ticket. Verify thresholds against your SLO window before recording.
- Tune for precision (few false pages) and recall (real outages caught) using replay of past data.
Draft outline details
Demonstration
Alert table: severity | long window | short window | burn rate | action. Plus a sketch of a 3-minute blip versus a 40-minute outage and which rules fire.
Check your understanding
Why add a short window to a burn-rate alert?
Show answer
So the alert resets quickly and only fires while the problem is ongoing.
Practice
Apply this lesson with the module's lab, exercises and quiz.
Marking a lesson complete is optional and you can undo it any time. Your progress stays in this browser.