Glossary
Uptime
Uptime is the share of time a service was available and doing its job, expressed as a percentage of a measured window.
Uptime is the proportion of a measured period during which a service was available. It is written as a percentage — 99.9%, 99.99% — and it only means anything alongside two other facts: the window it was measured over, and what counted as available. A month of 99.9% and a year of 99.9% are very different promises, and a service that answers every request in nine seconds can be 100% up and completely unusable.
The uptime formula is available time divided by total time, times 100. Over a 30-day month there are 43,200 minutes, so a month with 30 minutes of recorded downtime works out as (43,200 - 30) / 43,200 = 99.93%. The same 30 minutes over a 365-day year of 525,600 minutes works out to 99.994% uptime, which is why an uptime figure quoted without its window cannot be compared with anything.
The number is produced by sampling, not by continuous observation. A monitor checks on an interval and each check produces a verdict; uptime is the pass rate over those checks. That means the measurement's resolution is the check interval. If a service is checked every five minutes, a two-minute outage may be recorded as five minutes, as zero, or as nothing at all depending on when the checks landed — which is the practical argument for a one-minute interval on anything customer-facing.
Uptime is also not the same as the SLA it is compared against. An SLA usually excludes scheduled maintenance and defines availability in its own words; your monitoring records what actually happened. Keeping the two separate is what lets you answer both "were we up" and "did we owe anyone a credit" without arguing about which number is real.
| Check interval | Checks per month | Can miss an outage shorter than | One failed check costs |
|---|---|---|---|
| Every minute | 43,200 | 1 minute | 0.0023% |
| Every 5 minutes | 8,640 | 5 minutes | 0.012% |
| Every 10 minutes | 4,320 | 10 minutes | 0.023% |
| Every 15 minutes | 2,880 | 15 minutes | 0.035% |
Questions
What does uptime mean?
Uptime means the share of a period a service was available, written as a percentage. 99.9% uptime over a 30-day month means the service was unavailable for no more than 43m 12s of it.
How do you calculate uptime?
Divide the time the service was available by the total time in the period and multiply by 100. Monitoring does it by counting checks: the share of checks that passed is the uptime, which is why the result is only as fine-grained as the check interval.
What is a good uptime percentage?
For most customer-facing services 99.9% is the common baseline, which allows 43m 12s of downtime in a 30-day month. 99.99% allows 4m 19s, which is less time than it takes a person to notice and respond by hand, so detection has to be automatic. Each nine beyond that costs more than the one before.
What is the difference between uptime and availability?
Uptime is usually the measurement: what the service actually achieved over a period. Availability is usually the target it is held to, stated in nines in an SLA or SLO. Many teams use the words interchangeably; the useful distinction is between what happened and what was promised.
Does scheduled maintenance count as downtime?
In the measurement, yes: if the service was unavailable, a check records it. In an SLA it usually does not, because most agreements exclude maintenance announced in advance. Keep both numbers, and say which one a report uses.
In Uptimely
Uptimely records the verdict of every check, so uptime over any window is measured from stored results rather than estimated from incidents. A failed check counts as downtime and appears on your status page even when it does not page anyone — declaring an incident is a separate decision with its own threshold.
Uptimely checks as often as every minute on every plan, the free one included, and records every check, so your uptime is measured rather than estimated.
Start monitoring free →Read next
Product feature
Part of the Uptimely monitoring glossary — plain definitions of the vocabulary uptime monitoring, incident response and status pages are discussed in. Something here read wrong to you? Email support@getuptimely.com.