Uptime

What it is

Uptime measures how much of a period a service was operational. It's often expressed in nines: 99.9% allows about 43 minutes of downtime a month, 99.99% about four.

Why it matters

Uptime is a headline reliability promise and a frequent SLAService Level Agreement (SLA)A contractual commitment on response or resolution times. term. But a service can be 'up' while failing for one customer, which is why support tickets often catch problems uptime checks miss.

Best practices

1. Measure from the outside – Synthetic checks see what customers see.
2. Define 'up' carefully – Include degraded performance, not just total outages.
3. Reconcile with tickets – If customers report failures during 'up' periods, the check is too shallow.

How Decimal helps

Decimal investigates the customer-specific failures that slip past uptime checks and traces them to the service or config behind them.