Question

[Barry Margolin]

In article <387D9C38.4513 at dsea.unipi.it>,
Alessandro Coppelli  <Alessandro.Coppelli at dsea.unipi.it> wrote:
>I'm the administrator of a C class network.
>
>( where actually I have one a DNS server ( new Bind P5 ) )
>
> 
>
>I need for efficiency and security , to split my
>network in four parts .
>( I buying one switching layer 3 multiport )
>In one of this part I would put the servers, and in the
>other three parts all the clients. 
>I thought about two different solution to
>obtain the same ( I hope ) result.
>In the first I want to split  the net in four
>subnet, in the second solution  I would use three net of a C class not
>routable for the clients and for the fourth, the actual class, I would
>put the servers.
>
>What I want to know is: 
>how the configutation change in the 
>DNS server in each of the two solution ?

In the first solution you don't have to do anything special in DNS.  You'll
have a single reverse zone file for the whole class C, and just put all the
hosts in it.  It doesn't matter that the network is physically subnetted,
DNS just cares about the addresses.

In the second solution, you should create additional reverse zones on your
server for the private address blocks.  As long as your machines use your
server in their resolver configurations, they'll be able to perform reverse
lookups of the non-routable addresses.

-- 
Barry Margolin, barmar at bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Please DON'T copy followups to me -- I'll assume it wasn't posted to the group.