Best practice for naming computers

pc-rebuilders

Member
Reaction score
6
Location
Daytona Beach, Fl. US
If you have several computers, all the same model & operating sys, what is the best way to name them? I have a job with several and customer wants all of them back to factory settings. They are not running a server, but don't want to make a mistake. Cable modem and router...all of what they do is web based. I usually name them "Owner" -that being said, should I name them Owner1, Owner2 etc? or does it matter?
 
Family computers, whoever the primary person is (jon, jenny, sarah, doug, etc). If its a family computer, surname then.

Comapny computers, company name or company abbreviation, and then some type of designation number. For offices, office number. I.e. TARGUS113-PC
 
I used to do "Smith-PC" based on the family name but I've just started doing it "Smith-XPPro", "Smith-7HP" or "Smith-8Pro" so when they show up in GFI or ScreenConnect, I've got some extra instant information.
 
For company machines, I do pretty much the same thing, though I might put in a designation for location.

[CompanyName/Abreviation] [Location] [Equipment] [Number]

If you have a large number to manage, get some labels and label each with the name and serial.

So long as you are consistent, and the name has some rhyme and reason.
 
Putertutor has it dead on. Same with xander, though I'm not worried about os version in the name, but I see where he is coming from
 
If I named "Owner" at the start, can I change it to something else. if I remember right, whatever you name a box, you're stuck with it.

You can rename computers no problem...a quick couple of clicks and strokes of the keys, a reboot, done!

As for the question "best practice"...don't really have one, I leave it flexible based on the client. I don't care about homes/residential...supposed "surnames" is fine for that. But we do business networks. Servers...I often come up with a trend based on the client. Some abbreviation of the clients name perhaps, with a dash and the function of the server. "Client-DC" for domain controller, "Client-Exch" for Exchange Server, etc. Or I'll come up with funny names....like "Headnurse" for a home healthcare agency, or "Tanker1" for an oil delivery company.

As for workstations, I commonly do the department/position name. "Finance", or "AcctPay" for Accounts Payable, "FrontDesk", "NursingSuper", "Marketing", "Facilities", etc.
 
Names

I guess what I'm referring to is on a factory restore, if I name the computer "Owner", and change it later as per previous suggestions, when you click the start button, you still see "Owner" - Renamed one AMS01, worked fine, but seeing owner as mentioned. Or does it not matter?
 
You're seeing the username, not the computer name. If you want them to match, then rename the username too.

I usually do a few digits for the Clients name, then I have a simple letter based designation for what type of system it is. A server almost always has server in the name.

Workstations: SH-A01

SH for company name, A is a workstation, 01 is the first one I named.

Then I try to also slap a simple white sticker on the side of it, at least during deployment, and keep a simple notepad list on the server of the computer name with the location next to it.
 
I guess what I'm referring to is on a factory restore, if I name the computer "Owner", and change it later as per previous suggestions, when you click the start button, you still see "Owner" - Renamed one AMS01, worked fine, but seeing owner as mentioned. Or does it not matter?

You're confusing the PC's network name and the user's name.

We assumed your original post was about the PC's network name and the responses have been relative to that. The network name is easy to change and the only requirement is that it be unique within your network.

However, you last post talked about the name that shows when you click the Start button: that's the name of the userid to which you're logged on. There are actually two userid names: one is the name of the profile and is set when you create the userid, and the other is the display name (which is what is displayed when you click Start).

Changing the displayed name is simple: control panel -> user accounts, whereas changing the profile name generally requires creating a whole new profile. Many people create a profile name of OWNER and have the display name reflect the person/location.

p.s. the above is for a non-domain environment.
 
Best Practices
  • Make them unique
  • Do NOT include the User's Info on Business Machines (because users often switch computers)
  • Make it something that should never change.
  • Do NOT store duplicate info like Asset Tag, Organization Name, or OS Name (instead de-duplicate)
  • We just use the Service Tag or Serial Number


I know you said they have no AD, but the OS, Service Pack, etc. will ALL re-populate in AD for any large organization and it ensures you don't have duplicate names and that they don't get messed up.

i.e. You will ONLY have one 3UGLMB1 in an entire organization, so it is perfectly safe, where if you do something like the username PEGGYSUE-30129 or whatever the asset tag is, pretty soon you have depreciated objects like TINARAY-30129 etc.

... and you find yourself running a search like *-30129 and getting like 5 results... then you realize the stupidity of your customer.

In contrast, if someone later re-joins 3UGLMB1 to the domain, that object will get found and re-linked. It also prevents DNS issues etc.

The correct place for an asset tag is in the SMBIOS, so inventory software can query it.
 
Servers, just pick a naming convention and stick with it. Greek Gods, Pokemon, Their purpose (MailServer102 for rack 10 slot 2), but leave it documented and not too silly. I say Pokemon because I've walked in on a job and come across servers like that. Mu was their main server DHCP / DNS / etc.

Workstations, I usually just use office numbers or position (Finance1023)

Personal computers at a small home? I let the random name generator take care of it.
 
Make them meaningful is the best advice I can give you.

You want to easily and quickly know exactly which PC that network name is referring to.

I suggest somewhat shorter names, concise and meaningful.


Example:

Data Server: TN_DATA
Exchange: TN_EXCHANGE
Workstations: TN_WS101 first workstation on first floor
TN_WS201 first workstation on second floor

Somewhat of a rudimentary example, but those are good examples of unique, meaningful and static network names. They should never need to change, even if the user of the workstation (or the workstation itself) needs to be changed.

As far as user names are concerned, just make them meaningful and unique. Throwing some "extra" information in them can be helpful per the reasons others have mentioned.
 
One argument for using at least some user information in the machine name is for RWW. If users have to pick their machine from a list, you get fewer mistakes (=calls in the evening and weekends) if the names are obvious. I use initials at the beginning and the rest of whatever convention you use after that. When folks leave or move around, just rename the machine.
 
In my ethical hacking courses, we were taught well named servers are the best. If I gain access to a network, and I am hunting for a particular file, and they have 20 servers, it can be a pain. However, if they have them named by department, or whatever, it now makes life easier. same thing with desktops. If I know the researchers name, and the computer has his name on it, well that made life easy.

One suggestion we were told is instead of room number, department, or person, use the port number. Or the serial number of the machine. Having a good network map would greatly assist in this.

Now that I am not doing security work, am I always going to do what $20,000 in training taught me? No, lol. However, depending on the client and the risks, you should not assist a hacker. You cannot stop them, but why make it easy? Make them work for what they want.
 
Last edited:
^^^ yeah enterprise places with tons of servers usually have obscure names for all those racks in server cabinets. Outside the scope of probably everyone in this forum though.
 
^^^ yeah enterprise places with tons of servers usually have obscure names for all those racks in server cabinets. Outside the scope of probably everyone in this forum though.

For our internal stuff we decided to name each server after a Greek god.
At my ft job servers are named based on task and desktops are company initials then 3 numbers laptops have an L in their name.
 
^^^ yeah enterprise places with tons of servers usually have obscure names for all those racks in server cabinets. Outside the scope of probably everyone in this forum though.

And I think in that same train of thought that most of the clients who are serviced by most of those in this thread have no real need to worry about it.

It's not like their multi billion dollar business types that would have the kind of unwanted attention to need to take it that far.

At least I think so anyways.
 
Once a breach has happened there is really nothing you can do other than clean up. So I see naming as a way of trying to make things a bit harder. I will keep the naming somewhat consistent but will try to select names that do not openly give away the functions.

But to be honest even the biggies don't always try too hard to hide the function via the name. Was at a large bank site this evening. Looking over the names in the racks gave me a pretty good idea of what the functions were of the various servers. Many others sites regularly use things like location/store ID's in the server names. Especially retail.

I think that many large sites rely on the physical access control angle so they do not worry to much about the naming convention for security.
 
Back
Top