The emergency call comes in at 3:00 A.M., when your pager goes off with a message from your boss that a vital part of your network is down. You realize that one of the services has failed, and the server needs to be rebooted immediately. There’s only one problem. You don’t have a clue which server that vital service is located on. There’s a list of what each server has on it, but it’s located on a machine that may or may not be connected to the network at this point.

Obviously, this is a worst-case scenario. However, there’s one solution that could have made the situation much easier to troubleshoot: having a naming convention in place for the hardware on the network. Before your next disaster, it would be a good idea to come up with a naming convention and get everyone in your IT department to start using it so that all new equipment is named with the convention in mind.

You’ll want to extend the convention to all equipment that can have a host name (or NetBIOS name) on your network—servers, workstations, routers, switches, etc. For the purposes of this article, we’re going to look at how to develop a naming convention for your servers and workstations. You can take the general principles we cover here and extend them to naming other devices on you network as well.

What’s in a name?
The first thing you will need to consider is that a naming convention is not a set of nicknames for your servers. Naming each machine after a member of the cast of South Park might be cute and memorable, but it won’t help you or your associates who may be trying to figure out exactly what “Cartman” does during the day. What you call your servers when taking their names in vain is one thing but make sure you give them official names based on a solid naming strategy—and stick to it.

Naming conventions can be very simple or very complex, depending on your organization. For small offices, you may simply want to stick with short, descriptive names for your assets. For example, call each server what it does (mail, file, backup, etc). Call each workstation and laptop by its owner’s first initial and last name or network username. If your organization has a high staff turnover rate, you may simply want to give your workstations a number or a name based on physical location or department.

Bigger and better
Larger networks require a bit more planning before you decide on a naming strategy. Here are some questions to ask:

  • How many locations do you have to cover?
  • How many physical servers are there?
  • How should you handle workstations and laptops?

There are many other factors, but knowing the physical size and scope of the machines you have to name is a good place to start.

Next, determine what each server does. This is vital in naming the servers; you want to be as descriptive as possible. In our earlier example, if the service that failed shared its name with the server, finding which one to reboot would be a no-brainer. Once you have your lists of physical assets and what they run, you can begin to plan your naming strategy.

As an example, let’s take our old friends at Generic Networking Inc. GNI has offices in two cities, New York and San Francisco. At each office they have three servers. One handles backup and restore functions, one is an e-mail server, and one is a file and print server. Each office also has at least 15 client machines—both workstations and laptops. The lead engineer has decided it’s about time to name all these machines properly. Now, keep in mind that some services, like Microsoft Exchange, are dependent on the machine name they are installed with, so be very careful when changing machine names. GNI needs to upgrade its Exchange systems, so now is the perfect time to change names.

The lead engineer decides on the following convention for servers:

This translates to the two-character postal abbreviation for the city where the server resides, the abbreviated name of the main service on that machine, and a sequential number starting with 01; all separated by underscores. So the first Exchange server installed in New York would read:

The abbreviations for service names are arbitrary and should be decided upon and known by the technical staff of your organization so that there is no confusion. If multiple services run on a single machine, the process that is most vital on that machine should win out in the name.

This takes care of the servers, so GNI then went to work on its client machines. Our lead engineer agreed that first initial, last name is a good convention, but that doesn’t tell him anything about the machine’s type or location. So he decided on the following convention for client names:
City_FLastname_(M or W)_OS

This translates to the postal abbreviation for the office location the user is attached to, the first initial and last name of the main user (or general account name if it is a multiuser machine), a single letter to designate Mobile computer or Workstation, and an abbreviation for the operating system. According to this method; Peter Public, working in San Francisco on a Windows 2000 Pro laptop, would see his machine name as:

When the time comes to update software, GNI’s tech shop will know just by looking at the Network Neighborhood list who should get the updates based on username and OS. And if a user isn’t there, they will know by looking up the machine name if the user is mobile or just shut off his or her workstation.

All the info, right at hand
As you can see, a proper naming convention can save a lot of headaches down the road. Not only will you be able to immediately identify what servers are used for what purpose, but where they are and how many servers you’ll find there serving that purpose. Proper naming conventions on client machines will tell you who has the box, where they are, and what they’re running—which can make upgrades a breeze.

You will have to decide on a convention for your organization. This suggested strategy may not work in every company, but it is a good starting point for helping to determine how you will create your own conventions to help your organization now and for years to come. Don’t forget to document everything. A great naming strategy is going to work only if everyone on staff can figure out what the names mean.

So minimize confusion and maximize organization by developing an effective naming scheme. Then, the next time you get a call at 3:00 A.M., take some comfort in knowing you’ll be back in bed soon.

What kind of naming convention do you use?

We look forward to getting your input and hearing about your experiences regarding this topic. Join the discussion below or send the editor an e-mail.