Numbers That Matter
Uptime: the number, the measurement, and the gap between them
Every host publishes a percentage, and almost none of them publishes how, how often, or from where it was measured.
The short answer
Uptime is the proportion of time a service is genuinely reachable, and it only means something once you convert it into minutes. 99.9% allows about 43 minutes offline a month; 99.99% allows roughly four; 99% allows more than seven hours.
The measurement method decides the headline as much as the engineering does. Probe interval, probe location, whether planned maintenance is excluded and whether a slow response counts as a failure can move the published figure by a full nine.
Our target is 99.9%, monitored around the clock, and a month that falls short because of a fault on our side entitles you to a pro-rated credit under our terms.
By the HostingFast team · Reviewed 12 August 2026
100+
Terms defined here
2 min
Typical read time
Plain
English, no hand-waving
24/7
Engineers on shift
Start with the arithmetic. A 30-day month contains 43,200 minutes, so 99.9% permits about 43 of them and 99.99% permits about four. Those two figures describe very different engineering and very different price tags, which is why the number of nines is the only part of an availability claim worth arguing about.
Then look at how the figure is produced. A probe running every five minutes cannot see a 90-second outage at all. A probe inside the provider's own network measures the provider's network rather than the internet's view of it. Neither is dishonest, and neither is what your visitors experienced.
Keep your own record
Point an external monitor at your own domain, checking every minute from outside the host's network, and let it run. Within a month you own a log that nobody else controls, timestamped and independent, which is worth more in a support conversation than any dashboard the provider shows you.
Monitor something meaningful while you are at it. A 200 response on the home page can coexist with a broken checkout and a mail server refusing connections. Check a page that touches the database, and check the ports that matter to you.
Downtime costs more than the minutes
A visitor who meets a dead site rarely tries again, and crawlers record every failed fetch. Availability is the floor beneath every other metric you work on: a perfect Largest Contentful Paint is worth nothing during the 43 minutes the page did not answer at all.
It also compounds with speed. Sites tend to fail under load, which means outages arrive at exactly the hour your traffic is highest and your revenue is most concentrated. That is an argument for caching as much as for redundancy.
Reading an availability promise properly
Four questions settle it: what is the target, how is it measured, what happens in a month that misses it, and is there a status page with history still visible. A provider willing to answer all four in plain language is telling you something about how they operate.
Treat 100% claims with suspicion. Maintenance, network faults and hardware entropy make it impossible as a standing promise, and an unachievable number in the marketing is a reasonable predictor of what the support replies will be like.
What we target and what happens when we miss
99.9% is the figure we run to, monitored around the clock, and we deliberately call it a target rather than a contractual SLA. If a month falls short because of a fault on our side, our terms entitle you to a pro-rated credit — ask and it is applied. Hardware and network faults surface through platform monitoring, usually before the first customer notices.
The dedicated Xeon E-2274G is backed by a 99.9% uptime guarantee as part of its specification. Read Backup and CDN next, because availability and recoverability are separate problems that get confused constantly.
Availability, converted into minutes
Percentages are designed to be reassuring. Minutes are not, which is why this entry converts every figure before discussing it and then explains how the measurement itself can shift a headline.
Our own target is 99.9% with a pro-rated credit when a month misses it because of a fault on our side, which is a statement you can hold us to rather than a slogan.
- Nines converted into minutes a month
- Probe interval and location matter
- Keep an independent log
- 99.9% target with a pro-rated credit
Why HostingFast
Standard on every plan
A 99.9% target, monitored
Platform monitoring runs around the clock, and faults usually reach us before they reach your inbox.
A pro-rated credit when we miss
A month short of target through a fault on our side is credited under our terms — ask and it is applied.
Redundant power and carriers
The London datacentre runs backup power, industrial cooling and several independent upstream networks.
DDoS filtering at the edge
Volumetric traffic is absorbed before it reaches the machine serving your pages.
Daily backups behind the target
Availability and recoverability are different problems, so both are covered rather than one standing in for the other.
Engineers on shift at any hour
A one-hour first-reply target when a site is down, on a weekend and at 3am alike.
Quick Start
From order to online
- 1
Set up an external monitor
One-minute checks from outside the host's network, pointed at a page that touches the database rather than a static home page.
- 2
Convert every claim to minutes
99.9% is about 43 minutes a month, 99.99% about four. Compare hosts on the minutes and the difference stops being abstract.
- 3
Test a restore before you need one
Uptime keeps you serving; backups get you back. Restore a file into staging once so you know the second half works too.
Built In
Loaded onto every plan
- 99.9% uptime as the target, monitored around the clock
- A pro-rated credit when a month misses through our fault
- London datacentre with redundant power and multiple carriers
- DDoS filtering soaked up before it reaches the machine
- Daily backups with self-service restores from the panel
- Backups every six hours on the Nitro tier
- One-hour first-reply target when a site is down
- Two-hour first-reply target for everything else
- Imunify360 or ImunifyAV+ scanning every site
- Free SSL renewed automatically, so no expiry outage
Frequently Asked
What people ask us most often
How many minutes does each nine actually buy?
In a 30-day month of 43,200 minutes: 99% allows about 432 minutes, 99.9% about 43, and 99.99% about four. Each step is an order of magnitude of engineering and cost, so the honest comparison between hosts is the minutes rather than the marketing.
How do I verify a host's uptime claim myself?
Point an external monitoring service at your own domain, probing every minute from outside their network, and keep the log. It measures what your visitors received rather than what the provider's internal dashboard reports, and it is the only record you control in a support conversation.
Why do published uptime figures differ so much between hosts?
Mostly method. A five-minute probe interval cannot see a 90-second outage. Excluding planned maintenance removes scheduled downtime from the total. Probing from inside the provider's own network measures a shorter path than the one your visitors take. All three are legitimate choices that flatter the headline.
What happens if a month falls below the target?
Our terms provide a pro-rated credit where the shortfall was caused by a fault on our side; you ask, and it is applied. We call 99.9% a target rather than a contractual SLA on purpose, because a figure we consistently hit is worth more than a promise that would need lawyers to enforce.
Keep reading
Web Hosting
The layer that has to stay reachable, and the components deciding how quickly it answers when it is.
CDN (Content Delivery Network)
Edge caching that keeps pages serving to distant visitors even while the origin is under pressure.
How to Monitor Website Uptime
How to monitor website uptime yourself, step by step, including what to probe and how often.
WordPress Hosting
Managed WordPress with LiteSpeed and Redis, staging copies and daily backups included.
Web Hosting
Fast cPanel hosting on NVMe with SSL, migration and a first-year domain included.
Changing hosts? Run through our checklist first.
A straightforward sequence for a switch your visitors never feel: which files move first, how to shift email across without losing a single message, the right moment to repoint DNS, and the two mistakes behind almost all the downtime we get asked to rescue.
A target we hit, in writing.
99.9% monitored around the clock with a pro-rated credit when we miss — and engineers answering inside an hour when a site is down.
View WordPress Hosting plans