When your business depends on your website and your hosted applications, you can’t take any chances with a web host that isn’t going to spend the maximum amount of attention on the security of your data and the constant operation of your server. 99.9% uptime has long been an industry goal in the web hosting industry, but by applying many new backup and failover technologies, some web hosts are able to go beyond this level – achieving consistent levels of uptime that allow for only tiny amounts of disconnected time for upgrades and maintenance. What is your web host doing to make sure that your VPS keeps running?
Support and On-Call Engineers
The most important thing that your web host can provide to ensure that your web server runs smoothly is to provide some kind of constant and ongoing support. A good web server will give you a 24/7 phone number that you can call if there is downtime, immediately alerting an on-call engineer so someone can physically go to the data center to make sure everything is put back online as soon as possible. This is one of the biggest problems with low-cost shared hosting where the service is often non-existent. If something does happen to these budget servers, it can be days before someone answers your support ticket and your site is back online.
Backups are very important to ensure that your site and your VPS has continuity in operations. If something goes wrong, or your data is corrupted, a data backup will let you return your site to its last good condition without any interruption to your operations. Maintaining backups is thus a compromise between the desire to have data security and the need to fit your backups within the drive space that you are allocated for them. Your backups should never be stored on the same device as your server – since a single drive failure should never be able to delete your backup as well as your primary site. In many cases, web hosts will allow you to export copies of your backups so you can encrypt them and store them locally, or so you can upload them to the cloud host of your choice.
Unlike backups, data redundancy is designed to make sure that your website never goes down because of a drive failure. Hard drives have moving parts, and they gradually degrade until one day they either break or they stop working. Web servers should use some type of RAID array to ensure that your drives are constantly mirroring their contents on other drives – so one drive in the array can fail without wiping your data completely. Data redundancy can also take the form of copying your data over to multiple different physical systems – with the systems designed to come online if the primary system fails. If you have a high traffic website that has its own load balancing system, your data is already being stored on multiple servers, so when one server goes down then the other servers can temporarily pick up the slack.
Most data centers are reliant on one or two different internet connections – usually broadband connections – for their connectivity. This is very stable, and it provides high upload and download speeds, but it leaves the data centers vulnerable to internet outages. Sure, broadband usually has high uptime, but what if someone cuts a cable by accident, or the ISP has a major outage?
Internet connection failover is a relatively new industry – made possible by innovations in mobile networks. In most cases, you won’t need any type of connection failover – but if your site absolutely cannot go down, or you have critical server applications that you depend on to be up 100% of the time –it could be an option. For example, retail stores that rely on their internet connection during their busy seasons to allow for payments and checkouts usually use failover systems that transfer their internet connection to 4G and 3G networks whenever the internet goes out. Just in case the broadband doesn’t work – these devices switch over to a local mobile carrier to provide uninterrupted coverage without any real pause for the end user.