FM 15-01 Chapter 15 · Operations and capacity
SLA and uptime
Availability percentage to allowed downtime and back, composite availability and MTBF/MTTR.
Local only Runs in your browser. Nothing you enter leaves this page.
Availability to downtime
99.9 %3 nines
- Year (365.25 d)
- 8h 45m 57.6s
- Quarter (91.31 d)
- 2h 11m 29.4s
- Month, average (30.44 d)
- 43m 49.8s
- Week
- 10m 4.8s
- Day
- 1m 26.4s
Calendar months
- 28-day month
- 40m 19.2s
- 29-day month
- 41m 45.6s
- 30-day month
- 43m 12s
- 31-day month
- 44m 38.4s
A year here is 365.25 days (31,557,600 s), the average including leap years; quarter and average month are a quarter and a twelfth of it. An SLA measured per calendar month allows less downtime in February than in March, so check which the contract uses.
Downtime to availability
Accepts 4h 30m · 52 min · 1.5h · 01:30:00 · a bare number is minutes
- Achieved
- 99.45243 %
- Nines
- 2.26 nines
Composite availability
Tiers in series multiply. Inside a tier, N redundant instances give 1 - (1 - a)N. This assumes independent failures and instant, perfect failover, so real systems do worse.
- Load balancer
- 99.99 %
- Web server × 2
- 99.9975 %
- Database
- 99.95 %
- System
- 99.937506 %
- Downtime per year
- 5h 28m 41.4s
MTBF and MTTR
- Availability
- 99.954359 %
- Downtime per year
- 4h 3.3s
A = MTBF / (MTBF + MTTR). Steady state, averaged over many failures.