Uptime guarantee
Also called an availability guarantee.
An uptime guarantee is a provider's contractual commitment to keep a service available for a stated percentage of a measurement period. If availability falls short, the service-level agreement (SLA) sets out the customer's remedy, often credits against service fees.
How it works
The SLA defines the measurement period, what counts as an outage, exclusions and the claim deadline. Check whether "available" means the infrastructure is reachable or your application actually works. Credits can require a claim and supporting outage records; do not assume they are automatic.
What 99.9% allows in a 30-day month
A 30-day month contains 30 × 24 × 60 = 43,200 minutes. These calculations cover the full month with no exclusions.
| Uptime target | Calculation | Downtime at the threshold |
|---|---|---|
| 99.9% | 43,200 × 0.001 | 43.2 minutes |
| 99.99% | 43,200 × 0.0001 | 4.32 minutes |
A 60-minute qualifying outage misses both targets. Even then, the remedy may be a bill credit rather than reimbursement for lost sales.
Before promising uptime to customers
Questions people ask
What uptime does AWS guarantee?
There is no single AWS-wide guarantee. Under the Amazon Compute Service Level Agreement, EC2's Region-Level SLA is 99.99% for qualifying deployments across at least two Availability Zones in one region. Its Instance-Level SLA is 99.5%. Other AWS services have their own SLAs.
What does "ensure uptime" mean?
It means taking operational steps to keep a service running, such as redundancy and monitoring. The phrase alone does not specify a measurable commitment or a remedy.
