All levels of web-scale IT stacks should be monitored for availability, response time, and performance.  Whether self-imposed or customer-imposed, the availability SLA percentage translates to a specific amount of downtime.  Industry lingo may refer to this as the number of “nines” of an application or resource.  For example, 99.99 would be referred to as 4 nines (89% is 0 nines).

89% availability translates to the following duration of downtime: Daily: 2h 38m Weekly: 18h 28m Monthly: 3d 8h 21m Yearly: 40d 4h 14m
95% availability translates to the following duration of downtime: Daily: 1h 12m Weekly: 8h 24m Monthly: 1d 12h 31m Yearly: 18d 6h 17m
97% availability translates to the following duration of downtime: Daily: 43m Weekly: 5h 2m Monthly: 21h 55m Yearly: 10d 22h 58m
98% availability translates to the following duration of downtime: Daily: 29m Weekly: 3h 22m Monthly: 14h 37m Yearly: 7d 7h 19m
99% “Two Nines” availability translates to the following duration of downtime: Daily: 14m 24.0s Weekly: 1h 41m Monthly: 7h 18m Yearly: 3d 15h 39m
99.9% “Three Nines” availability translates to the following duration of downtime: Daily: 1m 26.4s Weekly: 10m Monthly: 44m Yearly: 8h 46m
99.99% “Four Nines” availability translates to the following duration of downtime: Daily: 8.6s Weekly: 1m Monthly: 4m Yearly: 53m
99.999% “Five Nines” availability translates to the following duration of downtime: Daily: 1s Weekly: 6.0s Monthly: 26.3s Yearly: 5m 15.6s
99.9999% “Six Nines” availability translates to the following duration of downtime: Daily: 0.1s Weekly: 0.6s Monthly: 2.6s Yearly: 32s