Burn-Rate Alert Playground

Tune SLO targets and multi-window multi-burn-rate alert parameters — the Google SRE way.
burn 0× OK

SLO & budget

The contract and the budget it produces.
Availability/success objective. Higher targets shrink the error budget hard (99.9% → 99.99% is a 10× smaller budget).
Rolling window the budget is measured over.
The current bad-event rate the simulation burns the budget at.

Burn-rate alert

Long + short window must BOTH trip to fire (Google SRE).
The slow window — decides if the alert is real.
The fast window — gates the alert so it clears quickly once the burn stops. Typically ~1/12 of the long window.
Multiplier over the “exactly spend the budget” rate. 14.4× = burn 2% of a 30-day budget in 1h.

Canonical presets

Snap to Google SRE Workbook pairs.
OK — not firing

Derived metrics

Recomputed live from your inputs.
Error budget
0.1%
allowed bad events
Burn rate
5.0×
errorRate ÷ budget
Time to exhaustion
6 d
at current burn rate
Budget left at exhaust
100%
of window allowance

Budget burn-down

Budget consumed over the window vs the 100% exhaustion line.
Budget consumed 100% (exhausted) Exhaustion marker

Generated PrometheusRule

Multi-window burn-rate alert reflecting current params.

Reference: canonical alert pairs

Google SRE Workbook, “Alerting on SLOs”. Budget consumed assumes a 30-day window.
SeverityBurn rateLong / ShortBudget burnedMeaning
PAGE14.4×1h / 5m2% in 1hFast, severe burn
PAGE6×6h / 30m5% in 6hSustained burn
TICKET3×1d / 2h10% in 1dSlow leak
TICKET1×3d / 6h10% in 3dChronic drift