Error Budget Calculator

How to use

1. 1Pick the SLO style: event-based (success ratio) when each request, query, or job counts as a good or bad event, or time-based (availability) when uptime in seconds is the metric.
2. 2Enter the SLO target as a percentage (99.9, 99.95, 99.99). Pick a window length: 24 hours, 7 days, 28 days, 30 days, or 90 days.
3. 3Type the share of the window already elapsed (50 for half a 30-day window, 100 for a window that just closed) so the budget can be denominated against the full window.
4. 4For event-based: paste the total events and bad events observed so far. For time-based: paste the observed downtime and pick a unit (seconds, minutes, hours, or days).
5. 5Read the burn rate, the share of budget consumed, the time to exhaustion at the current rate, and the multi-window alert table. Rows that would fire at your burn rate are highlighted in red.
6. 6Click Copy report to grab a plain-text summary for an incident review, an SLO policy doc, a release readiness checklist, or a status page footnote.

About this tool

Error Budget Calculator turns an SLO target, an observation window, and the numbers you already see in your dashboard into the four things SRE conversations actually need: the window's total error budget, how much of it has been spent, the current burn rate, and a projection of when the budget runs out. Two SLO styles are supported. Event-based SLOs measure the share of successful events; paste the total event count, the bad event count, and the share of the window that has elapsed, and the tool reports the event budget for the full window, the events already burned, the share consumed, and the time to exhaustion at the current rate. Time-based SLOs measure availability; paste an observed downtime (with a unit picker for seconds, minutes, hours, or days) and the tool reports the downtime budget for the window, the share burned, the projected exhaustion time, and the current achieved availability. The window picker covers the spans teams actually report on: 24 hours for intra-day alerts, 7 days for a weekly rolling SLO, 28 days (the Google SRE workbook default), 30 days for AWS billing-month basis, and 90 days for a quarter. A multi-window burn rate alert table renders the Google SRE workbook recommended thresholds (1 hour + 5 minute window at 14.4x, 6 hours + 30 minute window at 6x, 3 days + 6 hour window at 1x) and highlights in red any row that would be tripped at the current burn rate, so a team can decide which thresholds page on-call and which only file a ticket. A live progress bar shows the share of budget burned, color-coded against the SLO target. The math follows the standard error-budget definitions: budget equals (1 minus target) times window, burn rate equals observed error fraction divided by allowed error fraction, and exhaustion time equals remaining budget divided by the current consumption rate. A copy-ready report summarises every number for pasting into an incident review, an error budget policy doc, a postmortem, a quarterly SLO review, or a release readiness checklist. Useful for SRE teams running error budget policies, platform engineers writing freeze rules, on-call engineers deciding whether a burn rate is page-worthy, product managers tracking against an SLO commitment, and anyone managing a service against a measurable reliability target. All math runs in your browser, so traffic counts, error counts, and downtime values never leave your device.

Free to use. Works in your browser. No signup, no login.