Description of Windows 2003 Server DNS Stub Zones
DNS stub zones are a new feature in Windows 2003 server. What are they, what do they do and are they of any benefit? In this article I try and answer these questions and make the complicated subject of DNS that little bit easier.
2003 Server DNS Stub Zones Brief Overview
If you are already familiar with Windows DNS then you will be aware of secondary zones. Stub zones are similar to secondary zones in that they have a read only copy of the server that is authorative for a child DNS domain. The difference is that secondary zones contain a copy of all the A records from the authorative zone, stub zones contain just three records for each of the child domains, these are the SOA (Start of Authority) for the primary zone, the A record and NS record. This means that when a query is sent to the stub zone it can forward it straight to the correct DNS server as it knows it’s details.
How do Stub Zones WorkScenario A - Without stub zones being used. Scenario A - Without stub zones being used. Scenario A - Without stub zones being used. Scenario A - Without stub zones being used. Scenario A - Without stub zones being used.
Scenario A - Without stub zones being used.
Let’s take an example of how stub zones speed up the name resolution process compared to a system not using this technology.
Let’s say I’m sat at my PC and I want to find lisapc.thejkdomain.local. When this request is made a recursive query (recursive basically means it wants a full answer ie here is the address or that address doesn’t exist) is sent to my local DNS server (Server1). Server1 realises it’s not authorative for the requested info but it does know server2 is authorative for the .local domain so it goes off to server2 and asks for the resolution details. Although server2 is authorative for the .local domain it isn’t authorative for the thejkdomain.local domain but once again it knows who is. In this case server3 is authorative for thejkdomain.local. Server1 finally gets sent the requested info and can pass it back to me at my PC. SO to summarize
1.) Request sent from my PC to Server1.
2.) Server 1 passes info to server2.
3.) Then passed to server3.
4.) Sent back to server1.
5.) Server1 sends me the info.
Of course this all happens in super quick time anyway but wouldn’t it make sense to be able to cut down this time further. With stub zones we now can.
Scenario B - With stub zones being used
If we take the above scenario again but with stub zones being used then the procedure would be slightly different. I would still be requesting the info from server1. This time however because of the stub zone server1 would know that server3 was authorative for thejkdomain.local and would forward the request straight there. You can see by the number of steps that the total amount of steps is less.
1.) Request sent from my PC to server1.
2.) Server1 passes info straight to server3.
3.) Server3 passes info back to server1.
4.) Server1 sends me the info.
When to use stub Zones
However if you are in a one domain network then there is no benefit. Basically the more forests and domains there are the more reason there is to use stub zones. If you are setting up a multiple domain 2003 network then it makes sense to use them as a matter of course.
Why Use Stub Zones?
The idea behind stub zones is to speed up name resolution and reduce network traffic. This is a benefit for every network where you are able to use them.
By John Kellett
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 