# SLO & error budget calculator

Turn an availability target into the downtime and failed requests you can afford, and the burn-rate alerts that protect it.

## How it works

The error budget is the part of the window your service may fail: one minus the target, times the length of the window. A 99.9% target over 30 days leaves 0.1% of 43,200 minutes.

A burn rate of 1 spends the budget exactly over the window. Paging only when it burns many times faster, and only when a long and a short window agree, catches real outages quickly without waking anyone for a blip.

```
budget = (1 − target) × window
burn rate = error rate ÷ (1 − target)
```

## Questions

### What's the difference between an SLO and an SLA?

An SLO is the target you hold yourself to. An SLA is a promise to customers with consequences, usually set looser than the SLO.

### Which window should I use?

Rolling 28 or 30 days is common: long enough to smooth noise, short enough that a bad week still matters.

### What happens when the budget runs out?

Teams usually slow risky releases and spend time on reliability until the window recovers.

## Related tools

- [Burn-rate alert planner](https://powertraces.com/tools/burn-rate-alerts)
- [Uptime nines converter](https://powertraces.com/tools/uptime-nines)
- [Composite SLA calculator](https://powertraces.com/tools/composite-sla)
- [Latency percentile calculator](https://powertraces.com/tools/latency-percentiles)

[All tools](https://powertraces.com/tools)