High Availability Theory and the 9’s Measurement
High availability is an indicative measure of the “up time” of an application. The moment a host application/service stops servicing requests initiated by users or other application processes, the host application/service is termed to be unavailable and the “down time” clock starts ticking.
While negotiating availability with clients, non-functional requirement documents should clearly highlight the high availability quantifiers; Mean Time Between Failure (MTBF) and Mean Time To Recovery (MTTR).
Mean Time Between Failure (MTBF): Is calculated as (Hours / Failure Count) and is the average length of time the host application/service runs before failing.
Mean Time To Recovery (MTTR): Is calculated as (Repair Hours / Failure Count) and is the average length of time needed to repair and restore the host application/service after a failure.
Therefore, Availability = (MTBF / (MTBF + MTTR)) * 100
A popular way to describe high availability is by using the 9’s measurement:
-
Three nines (99.9 percent availability) represents about 8.5 hours of service outage in a single year.
-
Four nines (99.99 percent), represents about 1 hour of service outage in a year.
-
Five nines (99.999 percent) represents about 5 minutes of outage per year.
