Don’t Rely on DNS

In an increasingly IP-based world, DNS (Domain Name System) is a critical part of the foundation for communications. A device or web site is associated with an IP address, but it is much easier to remember to type ‘google.com’ into a Web browser than to try and remember that ‘google.com’ is actually located at 74.125.67.100.

Just trying to remember the links and sites you have bookmarked would be like trying to remember the combinations to open the lockers of all of your fellow students in your senior year (assuming you have either A) as many links bookmarked as I do, and/or B) a graduating class the size of mine). I have no idea what their locker combinations were, but I do recall the majority of the names of my graduating class peers.

So, DNS=good and DNS=important. What happend if we lose DNS then? Well, if the DNS server is unavailable for any reason you will not be able to visit web sites, connect with devices, or communicate with anything that is mapped or referenced base on its host name rather than its IP address.

That may not be a huge issue if users can’t access Google for a couple of hours, but losing DNS can also render VoIP and unified communications useless. That’s why its a good idea to build VoIP and unified communications infrastructures that map to the underlying IP address rather than the host name.

In situations where DNS is required, for example when NAT (network address translation) is used to map translated addresses to the proper endpoints, DNS should be deployed in a resilient fashion so that it does not become a single point of failure. The loss or lack of availability of one DNS server should not bring the unified communications system to its knees.