Fork me on GitHub
05 January 2013

MTBF, MTTD, MTTR and MTTF are the four parameters required to calculate the availability of a service or an individual component in a specific architecture.

  • MTBF is the Mean Time Between Faults
  • MTTD is the Mean Time To Detection
  • MTTR is the Mean Time To Repair
  • MTTF is the Mean Time To Failure

Well, once we got this, there are just a few little formula to know :

MTBF = MTTD + MTTR + MTTF
Availabiliy = MTBF / (MTBF + MTTR)

For example, a component with a MTBF of 8 days and a MTTR of 3 hours will have:
Availability = (8*24) / (8*24 + 3) = 0.985


A good way to understand availability is to look at the downtime per year they represent:

  • 90% availability = 36 days and 12 hours downtime/year
  • 99% availability = 87 hours, 36 minutes downtime/year
  • 99.9% availability = 8 hours,45 minutes, 36 seconds downtime/year
  • 99.99% availability = 52 minutes,33 seconds downtime/year
  • 99.999% availability = 5 minutes,15 seconds downtime / year
  • 99.9999% availability = 31 seconds downtime/year


blog comments powered by Disqus