SLO-Based Alerting Vocabulary
5 exercises — Practice the English for burn rate alert configuration, multi-window alerting, distinguishing fast vs. slow burn, and eliminating alert noise.
0 / 5 completed
1 / 5
A monitoring engineer configures an SLO alert and writes in the runbook: "This alert fires when the 1-hour burn rate exceeds 14.4×. Page the on-call immediately." What does this alert configuration detect?
14.4× is a significant threshold: the budget would be exhausted in ~2 days at this burn rate.
The math: monthly error budget ÷ 14.4 = 30 days ÷ 14.4 ≈ 2.08 days until total exhaustion.
Google SRE recommends this as a page-immediately threshold for services with a 99.9% SLO on a 30-day window.
Why 14.4×? It detects incidents significant enough to exhaust the budget within 2 days while still being specific enough to avoid false alarms for minor degradation.
Key vocabulary:
• Burn rate multiplier — 1× means the budget will last exactly the window; > 1× means faster consumption
• Page threshold — burn rate high enough to warrant waking up the on-call engineer
• Budget exhaustion time — window ÷ burn rate = days until the budget runs out
• Ticket threshold — lower burn rate that creates a ticket but doesn't page
The math: monthly error budget ÷ 14.4 = 30 days ÷ 14.4 ≈ 2.08 days until total exhaustion.
Google SRE recommends this as a page-immediately threshold for services with a 99.9% SLO on a 30-day window.
Why 14.4×? It detects incidents significant enough to exhaust the budget within 2 days while still being specific enough to avoid false alarms for minor degradation.
Key vocabulary:
• Burn rate multiplier — 1× means the budget will last exactly the window; > 1× means faster consumption
• Page threshold — burn rate high enough to warrant waking up the on-call engineer
• Budget exhaustion time — window ÷ burn rate = days until the budget runs out
• Ticket threshold — lower burn rate that creates a ticket but doesn't page