This article first appeared in issue 239 of .net magazine – the world's best-selling magazine for web designers and developers.
As sites become more complex, so does the hosting behind them. For even the most basic hosting account there are several key elements that need to be working perfectly harmoniously to keep your site online. This includes the datacentre, servers, network and the path from your provider over the internet to your computer.
If any of these elements has an issue – be that a website attack or something less sinister such as server failure - then your site would experience problems.
What constitutes ‘uptime’?
Scheduled maintenance is usually excluded from any uptime guarantees. Most providers will need to upgrade hardware and make changes, which can impact your site availability while the provider performs the work.
It’s also important to check if the agreement covers just the ‘network uptime’ or your actual site being online. The latter will usually only be offered on more complex hosting setups where the provider has more control. You would also need to understand how long an outage would need to be in order to get a service credit: seconds, minutes or hours?
How do the big companies do it?
Over the last 12 months, a multitude of issues have been reported on services that are provided by companies who have thrown millions at online offerings; Amazon Web Services (AWS), Microsoft’s cloud platform Azure, Blackberry and Facebook are just a few that had issues. However, there are a few simple ways to increase uptime and make yourself aware of issues before your clients realise.
How can I improve reliability?
Pick a hosting provider by looking at reviews, asking for uptime stats and even testing its support. If your site has an issue the provider is likely to be your first port of call so make sure it will be responsive and understand your needs. Ask: if you think your site will see bursts of traffic, can the provider handle it? If your site grows, what’s the next step? Planning ahead can help prevent downtime in the future.
Monitor your site. If you can be alerted to any problems before your users let you know, it’s not only more professional but lets you get to work on the problem and hopefully minimise any potential damage that arises from it.
Load test your site. It’s something that’s usually low down the list, but it can pay dividends. By load testing your site, even with a few visitors, you can identify any potential bottlenecks, which may cause downtime or slowness and put visitors off. It’ll also help you counteract the ‘TV effect’: how many times have you watched a company on Dragons’ Den only to visit its site and it’s down? Speak to your provider and it may be able to assist and manage the whole process for you.
Finally, it’s really important to have a backup of your site locally, or at least ensure your provider can supply you with one. Then, if your service has an extended outage, you have a recent copy of the site you can get online elsewhere.
Is 100 per cent uptime a myth?
So, you’ve taken precautions to prepare for service outages, but how far will a service provider’s guarantee go to protecting your site from downtime? Nothing is impossible. However, there’s a big difference between a few months of 100 per cent uptime and permanent 100 per cent uptime. Problems happen and, while ‘guarantees’ go some way to pushing a provider in the right direction, 100 per cent uptime is by no means a certainty.
The costs of providing a permanently online solution is still proving a challenge for even some of the world’s biggest companies. But, as technology evolves, it’ll become an easier battle to win. For now, having a hosting provider you can trust and work with, that can resolve issues quickly and point you in the right direction with site issues, may be more valuable than those few pounds you get back in a service credit.
How to send large files to clients: discover 8 free tools at Creative Bloq.