Glossary
Service level agreement (SLA)
An SLA is a contractual promise about service quality, such as 99.9% monthly uptime, with a stated consequence when it is missed.
Also called SLA, Service-level agreement.
SLA stands for service level agreement: a written commitment from a provider to a customer about the level of service it will deliver, usually uptime or response time, and what it owes the customer if it misses. A typical example: a hosting provider commits to 99.9% availability each calendar month and refunds part of the monthly fee when it falls short. Over a 30 days month, 99.9% allows 43m 12s of downtime; over a year it allows 8h 45m 36s.
The definition has two halves: a promise about service quality, and a defined consequence if it is broken. The consequence is what makes it an SLA rather than an aspiration — usually a service credit, sometimes a right to terminate. Everything else in the document exists to define precisely when that consequence is owed.
SLA time is the other kind of promise. Where an uptime SLA says how much of the month the service is available, an SLA time says how fast the provider acts: a response time (how soon someone acknowledges your ticket, for example within 1 hour) and a resolution time (how soon the problem is fixed, for example within 8 hours). The two are separate clocks, and "SLA time" on its own usually means whichever of them the contract defines. Averaged over many incidents, the resolution clock becomes MTTR, mean time to resolve, and the response clock becomes MTTA, mean time to acknowledge.
The definitions are where the substance is. An SLA states the measurement window (a calendar month, almost always), what counts as unavailable (often "all requests fail", not "some requests fail"), what is excluded (scheduled maintenance, force majeure, anything caused by the customer), and who measures. Vendors generally measure themselves. Read together, these clauses usually mean the promised number is easier to hit than a naive reading suggests.
Credits are also rarely automatic. Most SLAs require the customer to notice, to claim within a window, and to provide evidence — which is a good reason to measure your vendors independently rather than relying on their status page. It is also why an SLA is a commercial instrument and not an engineering target: the number you should actually run the service to is your SLO, and it should be stricter.
| Term | What it is | Example |
|---|---|---|
| SLA (service level agreement) | A contractual promise to a customer, with a credit or refund if it is missed | 99.9% monthly availability, or 10% of the fee back |
| SLO (service level objective) | The internal target the team runs the service to, stricter than the SLA | 99.95% of requests succeed over 30 days |
| SLI (service level indicator) | The measurement both are judged against | The share of checks or requests that succeeded |
| Availability | Downtime per month | Downtime per year |
|---|---|---|
| 99% (two nines) | 7h 12m | 3d 15h 36m |
| 99.5% | 3h 36m | 1d 19h 48m |
| 99.9% (three nines) | 43m 12s | 8h 45m 36s |
| 99.95% | 21m 36s | 4h 22m 48s |
| 99.99% (four nines) | 4m 19s | 52m 34s |
| 99.999% (five nines) | 26s | 5m 15s |
Questions
What does SLA stand for?
SLA stands for service level agreement: the part of a contract where a vendor promises a level of service, usually an availability percentage, and states what it owes the customer if that promise is broken.
What is an example of an SLA?
A common one is "99.9% availability per calendar month, measured by the vendor, excluding scheduled maintenance, with a 10% service credit when it is missed". Over a 30 days month, 99.9% allows 43m 12s of downtime before the credit is owed.
What is the difference between an SLA, an SLO and an SLI?
The SLI is the measurement, such as the share of successful checks. The SLO is the internal target for that measurement. The SLA is the external promise with a penalty attached. The SLO should be stricter than the SLA, so the team hears about a problem before a customer is owed anything.
What does 99.9% uptime mean?
The service is available for 99.9% of the measured period and may be down for the other 0.1%. That is 43m 12s in a 30 days month, or 8h 45m 36s in a year. A 99.99% SLA allows 4m 19s a month.
What is SLA time?
SLA time is the time limit an SLA sets for the provider to act. It is usually a response time (how soon someone acknowledges your request) or a resolution time (how soon the problem is fixed), and the two are measured separately. An SLA that promises uptime instead has no time limit, only a percentage.
What does SLA mean in simple terms?
It is a written promise about how good a service will be, plus what the provider owes you if it breaks that promise. "99.9% uptime or a credit on your bill" is an SLA.
Who measures whether an SLA was met?
Usually the vendor, from its own systems. That is why customers who expect to claim credits keep an independent record, such as external uptime checks, rather than relying on the vendor's own status page.
In Uptimely
Monitoring a vendor from your own account gives you the independent record an SLA claim needs. Uptimely records every status change with the time it happened and keeps each check's result for 30 days, so "we were down 41 minutes on the 14th" is a lookup rather than a recollection. Each monitor's Metrics page shows its uptime percentage for the last 7, 30 and 90 days, and a status page can show the same figure per component.
Uptimely checks your services and your vendors from outside and records every status change, so an SLA claim comes with its own evidence.
Start monitoring free →Related terms
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.