서버나 스토리지를 운영/선택 함에 서비스 운영 지속에 관한 지표가 있다.
9가 몇 개 인지에 따라 표현하는데, 아래와 같다.
|
Nines |
Unav. |
Availability |
Downtime per | |||
|
year |
month* |
week |
day | |||
| one nine | 10-1 | 90% | 36.5 d | 3 d | 16.8 h | 2.4 h |
| two nines | 10-2 | 99% | 3.65 d | 7.2 h | 1.68 h | 14.4 min |
| three nines | 10-3 | 99.9% | 8.76 h | 43.2 min | 10.08 min | 1.44 min |
| four nines | 10-4 | 99.99% | 52.56 min | 4.32 min | 60.48 s | 8.64 s |
| five nines | 10-5 | 99.999% | 5.256 min | 25.92 s | 6.048 s | 864 ms |
| six nines | 10-6 | 99.9999% | 31.536 s | 2.592 s | 604.8 ms | 86.4 ms |
| seven nines | 10-7 | 99.99999% | 3.1536 s | 259.2 ms | 60.48 ms | 8.64 ms |
| eight nines | 10-8 | 99.999999% | 315.36 ms | 25.92 ms | 6.048 s | 864 µs |
| nine nines | 10-9 | 99.9999999% | 31.536 ms | 2.592 ms | 604.8 µs | 86.4 µs |
* For monthly calculations, a 30-day month is used.
For large amounts of 9s, the "unavailability" index is easier to handle, which is also why such is used in e.g. hard disk bit error rates.
반응형
