How to use this calculator
- Enter or choose incidents: started, restored (one per line).
- The mean time updates as you type. Use Copy result or Share to keep it.
Method
Time to recoverᵢ = Restoredᵢ − Startedᵢ; MTTR = Σ Time to recoverᵢ ÷ n
- Startedᵢ
- when incident i began affecting service
- Restoredᵢ
- when service was fully restored
- n
- number of incidents
Worked example
- Incident 1: 08:00 → 10:30 = 150 min. Incident 2: 22:00 → 01:00 next day = 180 min. Incident 3: 60 min as entered.
- Mean = (150 + 180 + 60) ÷ 3 = 130 min = 2 h 10 min.
- Sorted 60, 150, 180, so the median is 150 min.
Across three incidents, mean time to recover is 2 h 10 min and the median is 2 h 30 min.
What the result means
Lower MTTR means shorter outages. Combined with mean time between failures it determines availability: Availability = MTBF ÷ (MTBF + MTTR).
Assumptions and limits
- Both timestamps on a line are in the same time zone; daylight-saving changes inside an interval are not adjusted.
- Partial degradations need a clear rule for when an incident “starts” and “ends”, or MTTR will not be comparable over time.
Questions
Is MTTR measured from the start or from detection?
For recovery it is normally from the start of customer impact, so detection delay is included. Some teams measure from detection; state which.
How does MTTR relate to availability?
Availability = MTBF ÷ (MTBF + MTTR). Halving MTTR has roughly the same effect on availability as doubling the time between failures.