Uptime Calculator
Calculate how much downtime your SLA allows, or find what uptime percentage you need for a specific downtime budget.
What is uptime?
Uptime is the percentage of time a service is operational and accessible. It's the primary metric used in Service Level Agreements (SLAs) to define reliability commitments between service providers and their customers.
When someone says their service has "99.9% uptime," they mean that out of any given time period, the service is expected to be unavailable for no more than 0.1% of that time. Over a 30-day month, that works out to about 43 minutes of allowed downtime.
Uptime is measured externally — from the perspective of your users, not your internal infrastructure. A server that thinks it's running but can't serve requests is effectively down.
Understanding SLA levels
Each additional "nine" of availability reduces your allowed downtime by 10x and typically requires a significant increase in infrastructure complexity and cost.
| SLA | Per Day | Per Week | Per Month | Per Year |
|---|---|---|---|---|
| 99% | 14m 24s | 1h 40m 48s | 7h 18m 18s | 3d 15h 39m 29s |
| 99.5% | 7m 12s | 50m 24s | 3h 39m 8s | 1d 19h 49m 45s |
| 99.9% | 1m 26s | 10m 5s | 43m 50s | 8h 45m 57s |
| 99.95% | 43s | 5m 2s | 21m 55s | 4h 22m 58s |
| 99.99% | 8.6s | 1m 0s | 4m 23s | 52m 36s |
| 99.999% | 0.9s | 6s | 26s | 5m 15s |
How to improve uptime
Use external monitoring
Internal health checks can't catch network-level failures or DNS issues. External monitors check from outside your infrastructure, the same way your users access your service. Check every 10-30 seconds from multiple regions.
Implement redundancy
Run at least two instances behind a load balancer. Use managed databases with automatic failover. Each single point of failure you remove gets you closer to the next nine.
Automate deployments
Manual deploys are error-prone and cause downtime. Use blue-green or canary deployments to roll out changes without taking the service offline. Automate rollbacks for failed deploys.
Set up alerting and on-call
Fast detection reduces downtime duration. Configure alerts with escalation policies so the right engineer gets paged immediately, with automatic escalation if they don't respond.
Run postmortems
Every incident is a learning opportunity. Document what happened, the root cause, and action items to prevent recurrence. The best way to improve uptime is to not repeat failures.
Frequently asked questions
How is uptime percentage calculated?
Uptime percentage = (Total time - Downtime) / Total time x 100. For example, a 30-day month has 43,200 minutes. If your service was down for 43 minutes, uptime = (43,200 - 43) / 43,200 x 100 = 99.9%.
What is the difference between 99.9% and 99.99% uptime?
99.9% uptime allows about 43 minutes of downtime per month, while 99.99% allows only about 4 minutes. The extra nine reduces your downtime budget by 10x, which typically requires significantly more infrastructure investment.
Does planned maintenance count against uptime?
It depends on how your SLA is defined. Some SLAs exclude scheduled maintenance windows from uptime calculations, while others count all downtime regardless of cause. Always check your SLA terms for the specific definition.
What SLA should I offer my customers?
Start by measuring your actual uptime over 3-6 months. Your SLA should be achievable based on real data, not aspirational. Most production SaaS services commit to 99.9% (three nines). Only commit to 99.99% or higher if your architecture genuinely supports it.
How do I monitor my actual uptime?
Use an external monitoring service that checks your endpoints from outside your infrastructure at regular intervals (every 10-60 seconds). Internal monitoring can miss failures that affect your users, so external checks from multiple regions give the most accurate picture.
Know your uptime. Protect your SLA.
PulseAPI monitors your endpoints every 10 seconds and alerts you the moment something goes wrong. Status pages, incident management, and on-call — all in one place.