MTTR, MTBF, MTTD
MTTR (Mean Time To Repair) measures time from detection to recovery — the number SLAs and postmortems obsess over. MTBF (Mean Time Between Failures) measures how often things break, lifted by better hardware and earlier intervention. MTTD (Mean Time To Detect) measures how long failures go unnoticed — the silent killer inside long MTTR. Availability = MTBF ÷ (MTBF + MTTR); move any lever and availability moves with it.
MTTR's biggest slice is finding, not fixing
When you break an incident into its timeline, the fix itself usually takes only minutes — swap a disk, restart, fail over. The hours are spent before that: noticing the failure, identifying which of five systems is actually in trouble, locating the device, and confirming what changed. So the biggest MTTR gains come from context, not speed: component-level early warning (driving MTTD toward zero), topology-correlated and business-impact-ranked alerts, an asset record that can answer "what is this and where is it," and instant out-of-band access.
FAQ
MTTR is Mean Time To Repair — the average time from failure to service recovery. MTTR = total downtime ÷ incident count, tracked separately by business and failure type. Ways to reduce MTTR: break down its components; component-level telemetry (faster detection); accurate asset and topology correlation (faster location); remote access and runbooks (faster repair). The bulk of MTTR hides in locating, not fixing.