Digital Lexicon Hub

The share of time a system or service is available and working as intended over a defined period, usually expressed as a percentage.

Every minute of unavailability can block sales, sign-ups and support, so uptime targets shape hosting choices, architecture spend and the remedies a company can claim from vendors under a Service-Level Agreement (SLA).

A 99.9% monthly target over a 30-day month (43,200 minutes) allows 43.2 minutes of downtime. If monitoring records 65 minutes, uptime is (43,200 - 65) / 43,200 = 99.85%, below the target.

Checks from a few locations can miss partial outages, slow pages or broken functions, and SLA definitions often exclude maintenance and very short incidents. A 100% target is neither realistic nor economically sensible for most services.

Time-based uptime is available time divided by total time in the period, multiplied by 100; cloud SLAs often define it as total minutes minus downtime minutes, divided by total minutes. Request-based availability instead divides successful requests by total requests. Targets are described in nines: 99.99% allows about 52.6 minutes of downtime per year. Monitoring tools probe endpoints at set intervals, and Observability data helps explain the failures behind the numbers.