MTBF, MTTR & Availability Calculator
Availability is how often something fails set against how fast you recover. Enter how long a component runs between failures and how long each recovery takes to get availability, yearly downtime, and how dependencies stack up.
See how chaining parts adds downtime
Downtime per year as more parts depend on each other
Each extra part in the chain must be working for the whole system to work, so downtime grows quickly. The highlighted row is the number of parts you entered.
Work backwards from a target
Pick the availability you need and see how reliable each part has to be — or how fast recovery has to get.
The formula
Availability = MTBF ÷ (MTBF + MTTR). Uptime over uptime plus downtime. Two levers move it: fail less often (raise MTBF) or recover faster (lower MTTR). They are symmetric in the maths — halving MTTR helps availability exactly as much as doubling MTBF — but recovery is usually the cheaper one to improve, through faster alerting, automated rollback and tested failover.
Dependencies compound
Components in series all have to be up at once, so their availabilities multiply. Ten dependencies at 99.9% each give a system at 0.99910 ≈ 99.0% — a tenfold increase in downtime. This is why teams count their critical-path dependencies and why "just add another service" is never free. Redundancy works the other way: two independent copies at availability A give 1 − (1 − A)², turning 99% into 99.99% — provided the failures really are independent and failover actually fires.
| Availability | Downtime / year | Downtime / month (30d) |
|---|---|---|
| 99% | 3d 15h 36m | 7h 12m |
| 99.9% | 8h 45m 36s | 43m 12s |
| 99.99% | 52m 34s | 4m 19s |
| 99.999% | 5m 15s | 26s |
Frequently asked questions
How do you calculate availability from MTBF and MTTR?
Availability = MTBF ÷ (MTBF + MTTR). If a component runs for 2,000 hours between failures and takes 2 hours to recover, availability is 2000 ÷ 2002 ≈ 99.9%.
What is the difference between MTBF, MTTF and MTTR?
MTBF (mean time between failures) is the average time from one failure to the next, for something you repair. MTTF (mean time to failure) is the average lifespan of something you replace. MTTR (mean time to recovery, or repair) is how long you are down each time. For high-availability systems MTBF and MTTF are close enough to use interchangeably here.
Why does adding dependencies make availability worse?
Components in series all have to be up for the system to be up, so their availabilities multiply. Five components at 99.9% each give a system at 0.999^5 ≈ 99.5%. Every dependency you add spends some of your reliability budget.
Is it better to fail less often or recover faster?
Both raise availability, but recovery is often the cheaper lever. Cutting MTTR in half does as much for availability as doubling MTBF, and faster detection, rollback and failover are usually more achievable than making a component fail half as often.
How does redundancy change the maths?
Two identical components in parallel (either one keeps you up) give availability 1 − (1 − A)². Two things at 99% become 99.99% together — as long as the failures are truly independent and failover actually works.