What is MTTR (mean time to repair)?

The average time it takes to repair a failed system or device and return it to a usable state (Mean Time To Repair). It includes the time to detect the failure, isolate the cause, replace parts or fix the settings, and verify the result. Availability is expressed as MTBF ÷ (MTBF + MTTR), so the shorter the MTTR, the higher the availability. It is shortened by monitoring that detects failures quickly, spare parts on hand, well-prepared procedures and automatic failover.

Japanese name: MTTR(平均修復時間)

MTTR (mean time to repair) appears in 4 practice questions across 3 topics, including IT Service Management and Systems Audit, Yield and quality control, Reliability.

Field: IT Service Management and Systems Audit

Calculating it from records

From the operating records of a period, MTTR is the total downtime divided by the number of failures. For example, if a system was down for a total of 6 hours in 3 failures during 720 hours, MTTR is 6 ÷ 3 = 2 hours, and MTBF is the operating time of 714 hours ÷ 3 = 238 hours. The availability is 238 ÷ (238 + 2) ≈ 0.992, which matches the fraction of time in operation, 714 ÷ 720.

Relation to availability

Since availability = MTBF ÷ (MTBF + MTTR), availability can be raised either by reducing failures to lengthen the MTBF or by speeding up repair to shorten the MTTR. For example, with an MTBF of 198 hours and an MTTR of 2 hours the availability is 0.99, but reducing the MTTR to 1 hour gives 198 ÷ 199 ≈ 0.995. Even when failures are hard to eliminate, the MTTR is relatively easy to shorten by improving monitoring and procedures, so it is an important target for meeting the availability in an SLA.

Ways to shorten the MTTR

Repair time can be divided into the time to notice the failure, to isolate the cause, to fix it and to confirm correct operation. Monitoring detects and reports failures immediately, configuration management information and logs quickly identify the affected equipment, spare parts or standby systems cut the time for replacement or switchover, and documented procedures let anyone carry out the work the same way. In a configuration that switches automatically to a standby system, the outage seen by users can be shortened greatly.

Try it: questions about MTTR (mean time to repair)

A system has an MTBF (mean time between failures) of 500 hours, and the availability target in its SLA is at least 99.5%. Assuming the MTBF does not change and calculating availability = MTBF ÷ (MTBF + MTTR), which of the following is closest to the maximum MTTR (mean time to repair) that meets the target?

  1. A1.5 hours
  2. B2.5 hours
  3. C3.6 hours
Show answer and explanation

Answer: B. 2.5 hours

Solving 500 ÷ (500 + MTTR) ≥ 0.995 gives MTTR ≤ 500 × 0.005 ÷ 0.995 ≈ 2.51 hours, so the maximum is about 2.5 hours. With 1.5 hours, the availability is 500 ÷ 501.5 ≈ 0.997, which meets the target, but it is not the maximum. 3.6 hours is the total outage time allowed in a 720-hour month (720 × 0.005), not the repair time per failure. Monitoring that detects failures quickly, spare parts on hand, well-prepared procedures and automatic switchover are effective in shortening the MTTR.

FAQ

What does the "R" in MTTR stand for?

It usually stands for repair, giving mean time to repair. In IT services it is also used in the sense of restore or recovery, the time until users can use the service again, so which interval is counted should be checked against the definition in the SLA or similar document.

How does MTTR differ from RTO (recovery time objective)?

MTTR is the average of the time actually taken for repairs, a value that describes past performance. RTO is the target for how long it may take to resume a business process or system after a disaster or failure, a value decided at the planning stage.

MTTR (mean time to repair) in the explanations

Questions that mention MTTR (mean time to repair) (4)

  1. A system has an MTBF (mean time between failures) of 500 hours, and the availability target in its SLA is at least 99.5%. Assuming the MTBF does not change and calculating availability = MTBF ÷ (MTBF + MTTR), which of the following is closest to the maximum MTTR (mean time to repair) that meets the target?
  2. Which process capability index is used when the process mean is offset from the center of the specification?
  3. Which correctly describes MTBF (mean time between failures)?
  4. During 30 days (720 hours) of operation, a system failed 3 times, and the total outage time due to the failures was 6 hours. Which of the following is the MTBF (mean time between failures) calculated from the results for this period?

Books on IT Service Management and Systems Audit Ad

As an Amazon Associate I earn from qualifying purchases.

Related terms

MTBF (mean time between failures) SLA (service level agreement) Process capability index (Cpk) MTTF (mean time to failure) Audit trail BCP (business continuity plan) CAAT (computer-assisted audit techniques) Change management Configuration management Escalation

Back to the glossary