Skip to content
Built in Lahore, PakistanHow we check the math
mathbehind

What 99.9% Uptime Really Means in Hours and Minutes

Each extra nine cuts allowed downtime by a factor of ten. Here is what three, four and five nines allow per year, month and day.

By Muhammad Ahmad, Software EngineerLinkedIn· figures checked

Published , updated . 4 min read.

Want to run your own numbers?Uptime and SLA Downtime CalculatorOpen the calculator

Uptime promises are written as percentages because they look impressive. 99.9% sounds close to perfect. It is easier to judge a promise once it is turned into time: how long the service can be down before the promise is broken.

The calculation

Subtract the uptime from 100% to get the share of time the service is allowed to be down, then multiply by the length of the period. Using a 365-day year:

Three nines: 99.9% uptime
  1. Allowed downtime share

    100% − 99.9%equals0.1%

  2. Per year

    0.1% × 365 d × 24 hequals8h 45m 36s

  3. Per month

    8h 45m 36s ÷ 12equals43m 48s

The nines at a glance

  • 99% (two nines): 3d 15h 36m a year, about 7h 18m a month.
  • 99.9% (three nines): 8h 45m 36s a year, 43m 48s a month.
  • 99.95%: 4h 22m 48s a year, 21m 54s a month.
  • 99.99% (four nines): 52m 34s a year, 4m 23s a month.
  • 99.999% (five nines): 5m 15s a year, about 26 seconds a month.

Going from three nines to four nines is not a small improvement. It means being down for under an hour a year instead of most of a working day, and it usually takes redundant infrastructure and on-call processes to achieve.

Your uptime depends on your dependencies

If your app needs a database and a payment provider, and both promise 99.9%, you cannot promise 99.9% yourself. If either one going down takes you down, the combined figure is roughly 0.999 × 0.999, or about 99.8%. Every hard dependency lowers the ceiling.

Read the fine print

  • Planned maintenance is often excluded from the calculation.
  • Some agreements measure per calendar month, which changes the exact minutes.
  • "Down" may only count full outages, not slow responses or partial failures.
  • The remedy for a missed target is usually a service credit, not a refund of your losses.

Which month? The period changes the minutes

Most service agreements measure uptime per calendar month, not per year. That sounds like a detail, but it changes how many minutes you are allowed. At 99.9%, the monthly budget is:

  • February (28 days): 40m 19s of downtime.
  • A 30-day month: 43m 12s.
  • A 31-day month: 44m 38s.

The "43m 48s a month" figure above is the yearly allowance divided by twelve, which is fine for planning. If you are checking whether a provider actually broke its promise, use the exact month the agreement names.

What an uptime promise is actually worth to you

When a provider misses its target, the usual remedy is a service credit: a percentage off that month's bill. It is worth doing the math on what that credit covers before you rely on it.

Suppose you pay $2,000 a month for a hosted service with a 99.9% monthly promise, and the agreement gives a 10% credit if uptime falls below that. A single three-hour outage in a 730-hour month brings uptime down to about 99.59%. You qualify for the credit, which is $200. If those three hours cost you $5,000 in lost sales and staff time, the credit covers 4% of the damage. An uptime promise tells you how seriously a provider takes reliability. It is not insurance, and your own plan for outages should assume you carry the cost.

Redundancy works the other way

Dependencies in a chain multiply their downtime, as the section above shows. Three hard dependencies at 99.9% each give roughly 99.7% combined, which allows about 26 hours of downtime a year instead of 8¾. Redundancy does the opposite. If you run two independent copies of a service and either one can handle the traffic, you are only down when both fail at the same time. Two copies at 99% each, failing independently, give 1 − (1% × 1%) = 99.99%.

The key word is independent. Two servers in the same data centre, on the same power supply, running the same faulty update, tend to fail together. Real redundancy means different zones, different regions or different providers, and it is the main way small teams reach four nines without heroic effort.

Using the number day to day: error budgets

Engineering teams often flip the uptime target around and treat the allowed downtime as a budget to spend. At 99.9% you have about 43 minutes a month. A deployment that causes a 15-minute outage has used roughly a third of the month's budget. If the budget is nearly gone, the team slows down risky changes until the next month. If it is barely touched, they can ship faster. It turns an abstract percentage into a practical rule for when to take risks.

Tip: Set your internal target a little tighter than the uptime you promise customers. It gives you room to respond before a breach.

Questions people ask

What uptime do cloud providers promise?
It varies by service and plan, often between 99.9% and 99.99% for core services, with credits if they miss. Check each service's own SLA.
How much downtime does 99.9% uptime allow?
About 8 hours 46 minutes a year, or 43 minutes 48 seconds a month.
How much downtime does 99.99% uptime allow?
About 52 minutes a year, or 4 minutes 23 seconds a month.
How do dependencies affect uptime?
Multiply the uptimes of hard dependencies. Two services at 99.9% each give roughly 99.8% combined.
How much downtime is 99.9% in a 30-day month?
43 minutes 12 seconds. In a 31-day month it is 44 minutes 38 seconds, and in February (28 days) 40 minutes 19 seconds.
What is an error budget?
The downtime your uptime target allows, treated as a budget. At 99.9% that is about 43 minutes a month; outages and risky changes spend it.