SLA Reference and Uptime Calculator

Uptime percentage tables, allowed downtime calculations, error budget guidance, and clear definitions for the three pillars of service reliability.

SLI vs SLO vs SLA

TermFull NameDefinitionExample
SLIService Level IndicatorA quantitative measure of a specific aspect of the service. This is what you measure.Proportion of requests completed in under 200ms
SLOService Level ObjectiveA target value or range for an SLI. This is what you aim for internally.99.9% of requests complete in under 200ms over a 30-day window
SLAService Level AgreementA contractual commitment with consequences for missing targets. This is what you promise externally.99.9% uptime; service credits issued for breaches exceeding 0.1% downtime

Best practice: Set your SLO stricter than your SLA. If your SLA promises 99.9%, aim for an SLO of 99.95%. The gap is your safety margin.

Uptime Percentage and Allowed Downtime

Uptime %Common NameDowntime per DayDowntime per MonthDowntime per Year
99%Two nines14 min 24 sec7 hr 18 min3 days 15 hr 40 min
99.5%--7 min 12 sec3 hr 39 min1 day 19 hr 50 min
99.9%Three nines1 min 26 sec43 min 50 sec8 hr 45 min 57 sec
99.95%Three and a half nines43 sec21 min 55 sec4 hr 22 min 58 sec
99.99%Four nines8.6 sec4 min 23 sec52 min 36 sec
99.999%Five nines0.86 sec26 sec5 min 15 sec

Error Budget Calculation

An error budget is the inverse of your SLO -- it quantifies how much unreliability your service can tolerate before breaching commitments. Error budgets enable data-driven decisions about feature velocity versus reliability investment.

Formula

Error Budget = 1 - SLO target

Example: SLO = 99.9%
Error Budget = 1 - 0.999 = 0.001 = 0.1%
In a 30-day month: 0.001 * 30 * 24 * 60 = 43.2 minutes of allowed downtime
SLOError BudgetMonthly Budget (minutes)Budget Burn Rate
99.9%0.1%43.2 minA 15-min outage consumes 34.7% of monthly budget
99.95%0.05%21.6 minA 15-min outage consumes 69.4% of monthly budget
99.99%0.01%4.32 minA 5-min outage consumes 100%+ of monthly budget

Practical Guidance

Choose the right number of nines

More nines cost exponentially more. Moving from 99.9% to 99.99% often requires redundant infrastructure, multi-region deployment, and sophisticated failover. Make sure the business value justifies the engineering investment.

Define "downtime" precisely

Is a 500ms response time "down"? Is partial degradation "down"? Your SLA should define exactly what constitutes a breach. Common approaches: error rate above threshold, latency above threshold, or complete unavailability.

Use error budgets to negotiate

When error budget is healthy, teams can take more risks with deployments. When error budget is low, freeze non-critical changes and focus on reliability. This creates an objective framework for the speed-vs-stability tradeoff.