No host is perfect, despite any claims they may make in their marketing materials. It is a fact of life that in any technology-based company, something can go wrong with the hosting server and a website can go down.
There are any number of reasons why there is hosting downtime. In many cases it can be external to the hosting provider. ISPs (eg Verizone, BT, Virgin etc) can have problems with their own network and this can mean that only sites hosted by a particular host are affected, based on location. This means that you may be able to access Twitter, Google etc but not your website. A common assumption when this happens is that your website is down everywhere. In reality, someone using a different, unaffected ISP will will have no problems at all. Such issues are not usually reported by hosts as it is not the host’s fault. Indeed, the hosts often don’t know about them unless informed. It is always worth checking to see if your site works on a mobile connection to see if the fault lies with the ISP, not the host.
Another common problem that will affect universal delivery of your website is hardware maintenance issues. Just like your personal computers need update and re-booting from time to time, so do hosting servers. It’s just a fact of technological life. When re-booting occurs a website will be down anything from a few minutes to a hour or so, while the server is re-staring and sorting itself out.
Longer downtimes can occur when there is a hardware issue and a part has to be replaced. Hosting companies have teams of software/hardware engineers ready to spring into action if there is a fault at the host’s data-center or with an individual server. The blips are often quickly repaired, but at other times, engineers may need to work for a few hours or even days to rectify a problem. This does not mean that your website will be down for days as hosting companies have backup systems in place, but it may mean that there is interruption to service which can last anything from a second or two, to a few minutes, to a few hours.
On other occasions, something freaky affects a server. In the same way that a PC can slow down or seize up for no apparent reason, so a hosting server can do the same. This can lead to short periods of downtime over a period hours until it’s resolved. Intermittent, freaky faults are the most difficult for hosting hosting engineers to track down and fix as it can take a long time to investigate and isolate the problem as the server behavior is unexpected and unpredictable. If a hosting company encounters such a fault, a status report is issued at the start and end of the problem.
There are a multitude of other reasons why a web server can fail. Often these result from the custom nature of a hosting providers system or the simple fact that hugely expensive server hardware systems can be affected by endless internal and external factors.
Customers are quick to point out occasional downtimes rather than giving credit to the amount of time is up. Most hosts are able to make claims of ‘99% uptime’. In reality, the 1% downtime is normal and represents any one of the host related issues above and is to be expected by any hosting provider. In their small print, no reputable host on the planet will guarantee 100% service and security. If they do, they likely to be dishonest, either intentionally or unintentionally!