What an Uptime SLA Actually Guarantees

Nearly every hosting provider advertises 99.9% uptime or better. That sameness hides real differences in what you're actually being promised.

Start with the math. 99.9% uptime allows about 43 minutes of downtime a month. 99.99% allows about 4. That tenth-of-a-percent gap between two providers' headline numbers is a 10x difference in permitted downtime, easy to miss while skimming a pricing page.

What the SLA actually covers matters more than the number itself. Most SLAs apply to specific services only, usually core compute, sometimes not managed databases or storage, and they carve out scheduled maintenance entirely. A provider can take your service down for a maintenance window and still hit their uptime target on paper. Read what's excluded before assuming the percentage covers everything you're running.

The remedy for a missed SLA is almost never money back in any meaningful sense. It's service credits, typically a slice of that month's bill, applied to a future invoice. A $20/month instance down for two hours might net you a 10% credit, around two dollars. That's not compensation for whatever the outage actually cost your business. Treat it as a modest apology, not insurance.

Getting the credit usually requires you to notice the outage, document it, and file a claim within a window, sometimes 30 days. Providers rarely credit you proactively. If uptime genuinely matters for what you're running, build your own monitoring instead of waiting for the provider to admit they missed their own promise.

None of this makes SLAs meaningless. They're a reasonable signal of how seriously a provider takes reliability, and the exclusions tell you something real about their posture. Just don't read "99.99% SLA" as "this will never go down." It isn't that, and the fine print says so.