There are thousands of web hosting companies. So how do you pick the right company to host your website, keep it online, and keep it secure? In this blog post, I’m going to talk about just 2 of the questions you should ask yourself.
- Does the provider offer an uptime guarantee?
Any shared web hosting provider should offer an uptime guarantee. An uptime guarantee is a promise that the servers will be online and available for a minimum amount of time within a given month. For example, a 99.9% uptime guarantee is a promise that your website won’t be down more than 0.01% in certain month (that percentage comes out to approximately 8 hours of downtime in a 30-day month). If a website is down for longer than that amount of time, then the guarantee essentially means the hosting provider will provide a refund or a credit in a specified amount.
Larger web hosting customers with dedicated servers may get something even a little bit stronger than a simple guarantee. A “SLA” (Service Level Agreement) is also a guarantee, but in the form of an executed contract between the provider and the client. While the uptime guarantee may not be enforceable in a court of law, a SLA is most certainly enforceable, assuming that there are no issues with the contract and both parties have signed the contract.
(Barred Owl Web does offer a SLA for our dedicated clients and an Uptime Guarantee for our shared clients)
- Does the provider maintain geographically redundant DNS servers?
In a workshop on “DNS & BIND” that I gave in 2013, I explained that DNS works transparently without any direct user interaction. A normal everyday user of the internet probably doesn’t even understand what DNS is and why it is so important.
DNS services are the backbone of the internet. It “tells” your internet browser and email client “where” a particular server is for a specific domain name. For example, the DNS servers that provide information on barredowlweb.com works behind the scenes to point “www.barredowlweb.com” to the server residing at the IP address of 18.104.22.168, and to point “mail.barredowlweb.com” to the server residing at the IP address of 22.214.171.124 (that is, as of the writing of this blog post).
If a DNS service isn’t working, very little that uses that provider’s DNS services are working either. Most network consultants will tell you that it is much worse for an entire domain name to go down due to DNS problems than it is for a website associated with that domain name to be down.
If a specific server goes down (such as the webserver at www.barredowlweb.com) and our other servers remain online, then if DNS still works then people would still be able to send email (which gets routed to mail.barredowlweb.com) and our clients would still be able to check their email.
If, on the other hand, our entire DNS infrastructure went down, none of these things would be possible – and worse, internet service providers would not even know where to route the requests to the website and email infrastructure to begin with!
So why is it imperative that a company not only maintain redundant DNS servers, but also make sure these DNS servers are geographically redundant (i.e. maintain 2 or more DNS servers that aren’t in the same geographical region)?
Due to possible power outages, natural & man-made disasters, and other scenarios that could affect an entire region, all of the servers within that region could become unavailable.