Unkillable Window

Website availability and downtime calculator

Calculate observed availability from a defined measurement window and recorded downtime. Compare it with a target without treating a target as a guarantee.

Availability = (window minutes − downtime minutes) ÷ window minutes × 100. Allowed downtime at a target = window minutes × (1 − target ÷ 100). A 30-day window contains 43,200 minutes; 30 minutes downtime means 99.93056% observed availability.

Illustrative starting values. Inputs stay in this tab. Download saves a plain-text record on your device; it includes any URLs you entered.

Source: https://unkillablewindow.dev/tools/availability/ · Save the assumptions with this result.

Before you enter your numbers

  1. Define the task, observation locations and start/end times.
  2. Merge overlapping outage intervals before totaling downtime.
  3. Use the same measurement window and exclusion policy when comparing results.

A worked comparison

Thirty days contain 43,200 minutes. A 99.9% target allows 43.2 minutes of downtime. Thirty minutes yields 99.93056% observed availability and is within that target. Sixty minutes yields 99.86111% and is outside it. The same outage in a 31-day window produces a different percentage.

Keep these assumptions with your result

  • Service or user task
  • Window and time zone
  • Monitor locations and interval
  • Non-overlapping outage intervals
  • Exclusions and partial failures
  • Evidence of recovery

Common questions

Does a 200 response mean the application was available?

Not necessarily. A login or checkout can fail while the homepage responds successfully. Define the user task your observation actually measures.

Does this calculate an SLA refund?

No. Contractual measurement, exclusions, claim windows and credit rules may differ. This tool only calculates the entered window and downtime.

How the tool works

Availability = (window minutes − downtime minutes) ÷ window minutes × 100. Allowed downtime at a target = window minutes × (1 − target ÷ 100). A 30-day window contains 43,200 minutes; 30 minutes downtime means 99.93056% observed availability.

What the result does not tell you

The tool assumes the downtime total is correct and non-overlapping. Probe failures, partial outages, planned maintenance and contractual exclusions need a defined measurement policy. A result does not establish an SLA credit or prove a service was available to every visitor.

Use the estimate responsibly

Replace the example values with records that cover the same period and scope. Save the input assumptions with your result so that another person can reproduce it. Change one assumption at a time to see why the result moves. Do not treat more decimal places as evidence that an estimate is more certain.

Keep working through the question