Voted Best Answer
Dec 11, 2018 - 01:59 AM
My number one piece of advice is KISS - Keep it Simple, S*****d.
My strong recommendation is that you simply use the asset tag number as the host name, perhaps prefexed with a letter so you know it is a host name.
If you don't include the asset tag number in your host name, then throw out your asset tags as they are effectively useless. They are also fairly useless if they can't be scanned with a simple barcode scanner, so if you don't already have bar code readable tags, now is your chance!!
If you don't have asset tags, use the serial number of the machine.
Your aim is to ensure that a physical piece of equipment sitting in front of an engineer is discoverable by its host name. That means the host name needs to include an identifier that is visually obtainable from the piece of equipment ie an asset tag or serial number. Adding extra bills and whistles to the host name simply makes that task harder and increases the administration they have to do when they are building the machine.
Your Ops guys may start arguing for complex codes that tell them what department uses the machine, if it's a laptop or desktop etc... but push back on those sorts of requests as all it does is complicate things.
You DO need all that information, but track it in the CMDB or Asset Register - so all the guys have to do is type in the asset tag and all the discovered and undiscovered bits of information about the machine will be in front of them.
Don't use different host names depending on device type (eg L123456 or D123456) - it just complicates the ordering of asset tags.
Don't use location or business unit or whatever as a basis - people move around and take their machines with them and it quickly becomes innaccurate.
Don't use finance or charge code information - people move around and it quickly becomes innaccurate.
As I said... KISS! Focus on making a physical piece of equipment quickly and easily discoverable on the network and in your CMDB and asset databases. That's it!