“Uptime” gets thrown around in marketing copy like it’s a single number that tells you everything. It isn’t. Knowing what it actually measures — and what it quietly leaves out — helps you read status pages and SLAs without being misled.
Uptime is the percentage of time a service was reachable and responding correctly over a given window. It’s measured by repeatedly checking the service and recording whether each check succeeded. Successful checks ÷ total checks = your uptime figure.
| Uptime | Allowed downtime / year |
|---|---|
| 99.9% | ≈ 8 hours 45 minutes |
| 99.99% | ≈ 52 minutes |
| 99.999% | ≈ 5 minutes |
Each extra “nine” cuts the allowed downtime by roughly 90% — and costs exponentially more to achieve. When a provider quotes a number, always ask: over what window, and what counted as “down”?
Most SLAs exclude scheduled maintenance, force-majeure events, and problems caused by the customer’s own code or configuration. The headline number can look great while real users still see outages. The exclusions are usually fair — but know them when comparing providers.
Check once an hour and a 5-minute outage can fall entirely between two checks — your log shows 100% while real users were affected. Shorter intervals catch more but cost more; most sites find every 1–5 minutes a reasonable balance.