Tips on ensuring high availability of your website.Posted on 25 March 2016 - Hosting
Web hosting is no different than any other system in that, in order for it to be as efficient, cost-effective and useful as possible, you want it to be available as much as possible.
As the old saying goes, time is money. And server downtime might mean no money at all is coming in during the time your website is unavailable. So, how can you organisation achieve high availability hosting for your website?
The first step in securing high availability hosting is to choose a web host that will guarantee a high level of availability in the first place. You might have a guaranteed level of availability written into a service level agreement (SLA) or you might be able to ascertain the availability by the infrastructure the host in question has in place.
For instance, here at Tibus, we only host in Tier 3 or Tier 4 classified data centres, which guarantee availability of 99.982% and 99.995%, respectively. That is obviously a nice starting point for achieving high availability hosting.
The next stage in improving hosting availability is to opt for a dual or multi-server hosting architecture. This means you do not just have a single server, but instead can rely on more than one. There are two main benefits of this in terms of availability.
Firstly, having more than one server mitigates against many of the problems that might cause downtime in a single server. For instance, if maintenance or a server reboot are required, your other server or servers can maintain website availability while this is being carried out.
Secondly, multiple servers allows traffic and load to be shared between the servers. This creates greater capacity and reduces the chances of your website become unavailable due to a sudden spike in traffic.
With all the right hardware in place, now you need a helping hand from some software. At Tibus, we use VMware’s vSphere software, which detects server and operating system failures and restarts the server when it detects any such failures.
Automating the restart process in this way further minimises downtime and helps you towards high availability hosting.