Uptime
Uptime is the percentage of time a service is available and responding correctly over a given period.
Uptime is the share of a time window during which a service is up and responding correctly. It is usually expressed as a percentage — 99.9%, 99.99% — and measured by an external monitor that checks the service on a schedule. Because the figure is a ratio, it always refers to a specific window: 99.9% over a rolling 30 days is a very different promise from 99.9% measured once a year.
Why it matters: uptime is the headline number customers, contracts, and status pages are judged on. It anchors SLAs and SLOs, drives service-credit math, and is the simplest signal of whether your reliability work is paying off. A single percentage point of difference can be the gap between minutes and hours of allowed downtime.
A common misconception is that uptime is an intrinsic property of your service. In practice it is a measurement, and its accuracy depends on how you observe it. A monitor that polls every five minutes can miss short outages entirely and will round detection lag into your numbers, so a generous-looking percentage can hide real customer pain. Checking more frequently, from multiple regions, produces an uptime figure you can actually defend.
Reliability targets are often described by their number of nines — three nines is 99.9%, four nines 99.99%. Each extra nine cuts the allowed downtime roughly tenfold, which is why each nine is dramatically harder and more expensive to reach than the last.
Uptime % = (Total time − Downtime) ÷ Total time × 100
Over a 30-day month (43,200 minutes), 30 minutes of downtime gives (43,200 − 30) ÷ 43,200 × 100 = 99.93% uptime.
Frequently asked questions
- What is a good uptime percentage?
- Most production web services aim for at least 99.9% (three nines), which allows about 8 hours 45 minutes of downtime per year. Customer-facing or revenue-critical systems often target 99.99% (four nines), under an hour per year.
- Does check frequency affect uptime accuracy?
- Yes. A monitor that checks every five minutes can miss short outages and inflates detection lag, so the reported number is coarse. More frequent checks — down to one second — produce an uptime figure that closely tracks what users actually experienced.
- Is uptime the same as performance?
- No. Uptime measures availability — whether the service responds correctly at all. A site can be 100% up yet painfully slow. Latency and error rate are tracked separately as service level indicators.
Start monitoring in under a minute
Free forever, no credit card. Upgrade when you need finer intervals.