IT Operations, Risk and Audit · #93

Online SLA Downtime and Availability Calculator

Shows how much downtime an SLA percentage really allows and whether current downtime exceeds the budget.

  • Data is processed in your browser and not sent to a server

Tool

Unplanned downtime in the period (minutes).

Dot decimals, separated by spaces (e.g. 99.9 99.95).

What does this tool do?

99.9 % sounds high but means 43 minutes a month and 8 hours 45 minutes a year. SLA negotiation, maintenance planning and post-incident reporting all need these minutes. The tool calculates the downtime budget per period, the actual availability from measured downtime and the total availability of a service built from dependent components.

How to use it

  1. Enter the target availability in percent.
  2. Optionally enter the period and total downtime in that period (minutes).
  3. Enter component availabilities separated by spaces (e.g. 99.9 99.95); press Run.

Fields

Target availability
Measurement period
Actual downtime
Unplanned downtime in the period (minutes).
Component availabilities
Dot decimals, separated by spaces (e.g. 99.9 99.95).

Worked example

InputValue
Target availability99.9
Measurement periodMonth (30 days)
Actual downtime30
Component availabilities99.9 99.95

Expected result: 99.9 %: downtime budget 43 min 12 s per month, 8 h 45 min 36 s per year. With 30 min downtime this month actual is 99.9306 %; 13 min 12 s budget left. Two components of 99.9 % and 99.95 % in series give 99.85 % in total.

Components in series (e.g. server + storage) make total availability lower than each; redundant (parallel) design raises it. Do not commit an SLA above the serial calculation of the components.

Second scenario

InputValue
Target availability99.99
Measurement periodYear

Expected result: 99.99 %: only 52 min 34 s per year, 8.6 s per day.

Method and assumptions

Downtime = period × (1 − availability). Month 30 days, year 365 days, quarter a quarter of a year. Actual = 1 − downtime / period.

Serial: A = A1 × A2 × …; parallel (redundant): A = 1 − (1 − A1) × (1 − A2) × …

Using the result and troubleshooting

State clearly in the SLA whether planned maintenance counts and what the measurement period is. Plan maintenance windows with the Maintenance Window tool and measure incident times with the MTTA/MTTR tool.

Limits and common mistakes

  • Partial outages (slowness, some users) and service credit calculations are not modelled; the parallel formula assumes independent failures.

Frequently asked questions

Why is a month 30 days?

It is the most common assumption in contracts; a 31-day month adds about 1.4 minutes of budget. Use the definition in your contract.

Related tools and service

Related service: Technical Support and Scheduled Maintenance

Systems online tools · See all online tools