Uptime SLA Calculator — what those nines actually permit
Convert an uptime percentage into the downtime it allows per day, week, month, quarter and year.
Free and instant — results appear in seconds. No sign-up, no limits, and nothing you type is stored.
Uptime percentages compress badly. The difference between 99% and 99.9% reads as trivial and represents an order of magnitude in permitted outage — roughly three and a half days a year against under nine hours.
This converts any percentage into real time across five periods, which is the form in which the number is actually decision-useful.
Worth checking against what you are being sold, and against what your own SLA promises if you resell.
The nines, in hours
99% → 7.2 hours/month 3.65 days/year
99.5% → 3.6 hours/month 1.83 days/year
99.9% → 43.8 minutes/month 8.77 hours/year
99.95% → 21.9 minutes/month 4.38 hours/year
99.99% → 4.38 minutes/month 52.6 minutes/year
99.999% → 26.3 seconds/month 5.26 minutes/yearEach additional nine divides permitted downtime by ten and multiplies the cost of achieving it by considerably more.
99.9% is the standard shared and VPS hosting guarantee. Forty-four minutes a month is one unplanned reboot and a slow recovery. Entirely achievable on a single well-run server.
99.99% requires redundancy. Fifty-three minutes a year does not permit a single server to fail and be replaced manually — you need failover that happens without a human.
99.999% is a data-centre-grade claim. Five minutes a year means no manual intervention is possible during an incident; recovery has to be automatic. Almost nobody outside large infrastructure operators genuinely delivers this, whatever the marketing says.
The question worth asking of any guarantee is not the number but the remedy. Most hosting SLAs credit a percentage of the monthly fee for the affected period, which on a small account is a rounding error against the revenue lost while the site was down. An SLA is a statement of intent with a token penalty, not insurance.
Read the exclusions
The headline percentage is the least interesting part of an SLA. What it excludes usually is:
Planned maintenance. Almost universally excluded. A provider can take your site down for four hours monthly with notice and still claim 100% uptime. Ask what notice period applies and whether there is a cap.
Measurement window. Monthly and annual measurement differ substantially. An annual 99.9% permits a single 8.7-hour outage with the guarantee intact; monthly measurement would breach on the same event.
What counts as down. Some providers measure a ping response, which stays healthy while your database is refusing connections and every visitor sees an error. Ask whether monitoring is HTTP-level.
Claim procedure. Credits are almost never automatic. Most require a written claim within a stated window — often 30 days — with your own evidence. Miss the window and you get nothing.
Cause exclusions. DDoS attacks, upstream network failures, and anything attributed to your own code are typically outside the guarantee.
Your uptime is the product of your dependencies
If your application sits on a 99.9% server, calling a 99.9% database and a 99.95% payment API, your realistic ceiling is the product of all three — roughly 99.75%, or around 22 hours a year, before you have written a single bug.
More dependencies means lower composite availability. This is why adding a third-party service is an availability decision as much as a feature decision.
Measuring your own
Provider dashboards report on provider infrastructure. That is not the same as whether your site was reachable and working.
External monitoring from several locations, checking an actual page rather than a ping, is the only figure worth arguing from. It also gives you the evidence a credit claim requires.
Monitor a page that exercises the database. A static homepage served from cache can look perfectly healthy while the application behind it is entirely broken.
Frequently asked questions
How much downtime does 99.9% allow?
About 43.8 minutes per month, or 8.77 hours per year. That is the standard shared and VPS hosting guarantee, and it is achievable on a single well-maintained server without redundancy.
Does planned maintenance count against the SLA?
Almost never. Nearly every hosting SLA excludes scheduled maintenance, so a provider can take your site offline for hours with notice and still report 100% uptime. Ask what notice period applies and whether the total is capped.
Is a 99.99% guarantee worth paying more for?
It depends on what an hour of downtime costs you. The step from 99.9% to 99.99% cuts annual downtime from about nine hours to under an hour, but it requires genuine redundancy and automatic failover, which is a substantial cost increase. For most small business sites, 99.9% with good backups is the better value.
What do I actually get if the SLA is breached?
Usually a credit against a percentage of the monthly fee for the affected period — rarely compensation for lost business. Credits are almost never automatic; most providers require a written claim within a set window, often 30 days, supported by your own evidence.
Why does my monitoring disagree with my host's status page?
Provider dashboards report on provider infrastructure, not on whether your specific site was reachable and functioning. External monitoring from several locations, checking a real page that exercises the database rather than a ping, is the figure worth arguing from.
Reviews
No reviews yet. If this tool solved something for you, yours would be the first — and it helps other people decide whether it is worth their time.