Uptime SLA Explained: 99.9% vs 99.99% Downtime Minutes

You sell hosting, VPS, or domains, and your SLA promises 99.9% uptime. But what does that actually mean when a server goes down at 3 AM? It means you have about 43 minutes of allowed downtime per month. Miss that, and you owe credits. This article breaks down uptime percentages into real minutes, explains what those windows mean for maintenance and monitoring, and shows you how to talk to customers when things go wrong.

What does 99.9% uptime actually mean in minutes?

99.9% uptime means your service can be down for up to 43 minutes and 12 seconds per month (based on a 30-day month). That's the maximum allowed downtime before you breach the SLA. For a 31-day month, it's about 44 minutes and 38 seconds. For a year, it's roughly 8 hours and 46 minutes.

Here's the math: 30 days × 24 hours × 60 minutes = 43,200 minutes. 0.1% of that is 43.2 minutes. So 99.9% leaves you 43.2 minutes of downtime.

How many minutes of downtime do different uptime SLAs allow?

Here's a quick reference for common uptime percentages, calculated for a 30-day month:

• 99% – 7 hours 12 minutes downtime
• 99.5% – 3 hours 36 minutes downtime
• 99.9% – 43 minutes 12 seconds downtime
• 99.95% – 21 minutes 36 seconds downtime
• 99.99% – 4 minutes 19 seconds downtime
• 99.999% – 26 seconds downtime

Notice how the allowed downtime shrinks dramatically as you add nines. Going from 99.9% to 99.99% cuts your monthly downtime budget from 43 minutes to just over 4 minutes. That's a tenfold reduction.

What do these windows mean for maintenance?

Maintenance is planned downtime, and most SLAs exclude it if you give advance notice. But you still need to fit maintenance into your uptime budget if it causes unplanned outages. For example, if you have a 99.9% SLA, you can't afford to take the server down for an hour every month for updates—that alone would blow your budget.

Instead, schedule maintenance during low-traffic windows and keep it short. Use rolling restarts or live migration where possible. If you must take a service offline, do it outside the SLA calculation by announcing it as planned maintenance.

How does uptime affect monitoring and alerting?

Monitoring is how you know if you're meeting your SLA. Set up checks that measure uptime from multiple locations. A single monitor can miss outages if the monitor itself fails or if there's a network issue on your end.

For a 99.99% SLA, you need monitoring that can detect outages within seconds, because your entire monthly budget is just over 4 minutes. If your monitoring only checks every 5 minutes, you could miss a 4-minute outage entirely and not even know you breached the SLA.

Use tools that track uptime percentage over time and alert you when you're approaching your downtime limit. That way you can prioritize fixes before you owe credits.

How should you communicate downtime to customers?

Customers care about two things: how long the outage lasted and what you're doing about it. Be transparent and proactive.

• Planned maintenance: Notify customers at least 24–48 hours in advance via email and a status page. Explain what you're doing, why, and the expected duration.
• Unplanned outages: Acknowledge the issue quickly on your status page and social media. Give updates every 30–60 minutes, even if there's no new information—silence makes customers think you're hiding something.
• Post-incident: After the outage, send a summary explaining the cause, the fix, and what you're doing to prevent it from happening again. If you owe SLA credits, process them automatically and tell customers.

If you use a billing and automation platform like Teculiar, you can integrate status updates and credit issuance directly into your client area, so customers see the impact and resolution without extra effort from you.

What should you promise in your SLA?

Don't promise 99.99% unless you can afford the infrastructure and monitoring to back it up. Most resellers start with 99.9% and move up as they grow. Your SLA should also define what counts as downtime (e.g., network, power, hardware) and what doesn't (e.g., scheduled maintenance, customer actions).

Be clear about credits: how customers claim them, how they're calculated, and any caps. A fair SLA builds trust; an unrealistic one leads to constant firefighting.

What to do next

• Calculate your current uptime using a monitoring tool and compare it to your SLA.
• Review your maintenance procedures to ensure planned downtime is excluded from SLA calculations.
• Set up multi-location monitoring with alerts for when you're close to your downtime limit.
• Create a communication plan for outages, including templates for status updates and credit notifications.

If you're looking for a platform that helps you automate billing, provisioning, and customer communication, see our pricing page for details.