For Windows Server Abstract

Naming Convention Guide

1. Presentation

1. 
   Server names must have a fixed length: In the current naming convention, we have different names lengths, meaning that we can’t easily use scripting to classify and automate tasks (sometimes, machine name is used in scripts to crosscheck operations in automated packages deployments). Example: If all machines in a test environment are starting with “T” then it’s easy to set a rule in the deployment console.
   
2. 
   Server location: In a today naming convention, the recommendation is not to include any location information in the server name. Indeed, with the arrival of virtualization technologies inside the datacenter, we can easily have a machine that is today located in a datacenter that moves to another location. In this case, the server location contained in the name will not be accurate anymore. If we would have the location information for each server, it is recommend to use one of the Active Directory field or another tool to put location information. Doing this is like treating the server location as a parameter or attribute that can be changed without impact. Indeed, changing a server name is technically feasible but will have consequences and for some products, it means full reinstallation. To align with this recommendation, no locations information have been included in this naming convention.
   
3. 
   Server name length: During discussions, we agreed that the server name should not be too long. As a best practice, the total server name can’t be more than 15 characters. Indeed, as soon as we are going above this limit, some software (like in Unix and Mainframe worlds) can’t communicate anymore with the Windows world as the server name is too long to enter in the field. The naming is therefore adapted to stay with the shortest name as possible.
   
4. 
   Special characters: Special Characters are generating issues when scripting is used. Indeed, in most scripting languages like PowerShell or Perl, special characters like “-“ are interpreted as “minus” signs and therefore scripts are failing. The best practice is to only use standard letters in a Windows Active Directory based server / desktop naming convention. All special characters have therefore been removed from the naming convention.
   

2. Intended audience

3. Sources

1. Change history

   Version	Nature of change	Date
   1.0	First version	20/09/2012

2. Forecast changes

   Version	Nature of change	Date

3. Abbreviations / Glossary

   Abbreviation	Full text

1. Naming Convention Structure

   1. The structure

• 
  Block 1: Provides the environment with one character
  
• 
  Block 2: Provides the Network Zone with one character
  
• 
  Block 3: Provides the Server role with one character
  
• 
  Block 4: Provides an application description view with three letters and two numbers
  
• 
  Block 5: Provides the service role provided by the machine
  

2. The naming in itself

“ENRAAAxxSSS” where we can see that each server has a name in 11 positions.

• 
  The “E” is representing the Environment.
  
• 
  The "N" is representing the Network Zone. These are zones that have been discussed and agreed with Network team.
  
• 
  The “R” is representing the server role. It’s describing, from a generic point of view, the role played by the server in the environment
  
• 
  The “AAAxx” are describing, for the three “AAA” the application that is running on top of the machine and the “xx” are replaced by numbers. These numbers must always start with “01” and increase when the number of machine is increasing. The current list of “Application descriptions” can be extended to cover new applications.
  
• 
  The “SSS” at the end of the machine name are describing Windows services that are associated with the server.
  

3. Naming Examples

• 
  1^st letter => P as it’s a production server
  
• 
  2^nd Letter => Database servers are in the backend zone => B
  
• 
  3^rd letter => Role is a database server => D
  
• 
  4^th to 8^th letter => SQL01 as this is the first SQL server (SQL is considered as consolidated here). The number here is indicating the cluster number.
  
• 
  9^th to 11^th letter => N01 for node one of the cluster and N02 for node two of the cluster. Numbers here are indicating the n° of each node inside the cluster.
  

• 
  Server names are:
  
  • 
    Node1 of the cluster: PBDSQL01N01
    
  • 
    Node2 of the cluster: PBDSQL01N02
    

• 
  1^st letter => P as it’s a production server
  
• 
  2^nd Letter => External zone => E
  
• 
  3^rd letter => Role is an IIS server => W
  
• 
  4^th to 8^th letter => ENJ01 as this is the first Enjaz server. The number here is indicating the IIS cluster number.
  
• 
  9^th to 11^th letter => L01 for node one of the cluster and L02 for node two of the cluster. Numbers here are indicating the n° of each node inside the cluster. It’s an “L” in this case and not an "N" as IIS nodes are load balanced, not part of an MSCS cluster.
  

• 
  Server names are:
  
  • 
    Node1 of the cluster: PEWENJ01L01
    
  • 
    Node2 of the cluster: PEWENJ01L02
    

• 
  1^st letter => T as it’s a test server
  
• 
  2^nd Letter => Application servers are in the Application zone => A
  
• 
  3^rd letter => Role is an hypervisor/virtualization server => V
  
• 
  4^th to 8^th letter => ESX01 as this node is part of the first ESX cluster.
  
• 
  9^th to 11^th letter => N03 for node three as this ESX host is the third ESX box being part of the first ESX cluster.
  

• 
  Server names is:
  
  • 
    TAVESX01N03