SLA Uptime Calculator
Enter an SLA percentage to instantly get the allowed downtime per year/month/week/day, or enter actual downtime to work backward to an uptime percentage — includes a standard SLA-tier reference table.
FreeOnline Tool
Loading…
How to Use
- In "SLA% → Allowed Downtime" mode, enter an SLA percentage (e.g. 99.9) and the tool instantly calculates the maximum allowed downtime per year, month, week, and day, along with the approximate number of "nines".
- Switch to "Downtime → Calculate Uptime" mode, enter the actual downtime that occurred (seconds/minutes/hours/days) and the reporting period (year/month/week/day), and the tool works backward to give you the equivalent uptime percentage, approximate nines, and equivalent yearly downtime.
- The "SLA Tier Reference Table" below lists downtime figures for 10 common SLA tiers from 90% to 99.9999%; your current input is automatically highlighted against the closest row, so you can see where you land.
- Click "Load Sample" in either mode (99.95% for SLA mode, 52.56 minutes/year for downtime mode) to see it in action immediately.
Features
- Bidirectional calculation: convert a known SLA percentage into allowed downtime, or convert known actual downtime into an uptime percentage — switch modes with one click.
- SLA mode returns allowed downtime for all four reporting periods (year/month/week/day) at once, plus the approximate number of "nines" (e.g. 99.9% ≈ three nines).
- Downtime mode automatically converts the result into equivalent yearly downtime, making it easy to compare against a contracted SLA.
- Includes a built-in reference table of 10 standard SLA tiers (90%, 95%, 99%, 99.5%, 99.9%, 99.95%, 99.99%, 99.995%, 99.999%, 99.9999%), auto-highlighting the row closest to your current input.
- All calculations run client-side (via logic.js) with instant feedback on input — no button click required.
Use Cases
Evaluating whether a vendor's SLA terms are reasonable
Before signing a cloud or colocation contract, plug in the vendor's promised SLA percentage to see exactly how much downtime per month/year it actually allows, then judge whether that fits your tolerance.
Calculating actual uptime after a postmortem
After tallying a quarter's cumulative downtime, an ops team switches to "Downtime" mode to derive the actual uptime percentage for the postmortem report and SLA compliance summary.
Setting internal availability targets
A team defining an internal microservice's availability goal references the SLA tier table to pick an appropriate number of nines, then sets alert thresholds and degradation plans accordingly.
Explaining what "nines" actually mean to non-technical stakeholders
A tech lead reports to a business audience by using the tool to show that "99.99% equals about 52 minutes of downtime a year" — a much more concrete number than a bare percentage.
FAQ
What does "number of nines" mean?
It's shorthand for how many consecutive 9s appear in an availability percentage — 99.9% is "three nines", 99.99% is "four nines". More nines means less allowed downtime and a higher availability requirement. The tool automatically calculates and displays this (as a one-decimal approximation).
Do the two calculation modes stay in sync?
Yes. After deriving an SLA percentage in "Downtime → Uptime" mode, switching back to "SLA% → Downtime" mode auto-fills that derived percentage — both modes share the same underlying calculation logic.
Is the monthly period based on 30 days?
Yes — the tool's month period is fixed at 30 days, the year period at 365 days, week at 7 days, and day at 24 hours. This is the common simplified convention used in SLA calculations industry-wide; check your actual contract's wording for the precise definition.
Why is a row always highlighted in the reference table even when my input isn't one of the listed values?
The table shows 10 fixed common tiers. The tool finds and highlights whichever tier is numerically closest to your current SLA input, so you can quickly gauge roughly which availability bracket you're in — it's a nearest-match, not an exact match.