SLO / SLI Calculator

SLI (Service Level Indicator) is a measured reliability signal, like the percentage of successful requests.

SLO (Service Level Objective) is your target for that signal over a time window, such as 99.9% over 30 days.

Error budget is the allowed failures before missing the SLO: total requests × (1 − SLO).

You can capture request counts and errors from logs, metrics, or tracing. The most common inputs for this calculator are total requests and error requests.

• Metrics (recommended): instrument your service to emit counters like http_requests_totaland http_requests_errors_total. Tools like Prometheus, CloudWatch, and Datadog can aggregate these over a time window.
• Logs: parse access logs (Nginx, ALB, CloudFront) and count total vs. 5xx/4xx responses. This is helpful when metrics are unavailable, but it is slower for large windows.
• Tracing: use trace data (OpenTelemetry, Jaeger) to compute success rates per route or dependency. This is best for deeper diagnostics but usually higher overhead.

Pick the same definition of "error" consistently (5xx only, 4xx + 5xx, timeouts, or custom business failures) to keep your SLI stable.