host declaration restrction (Re: ISC DHCP 3.0.5b1 has been released!)

[Glenn Satchell]

>Subject: Re: host declaration restrction (Re: ISC DHCP 3.0.5b1 has been 
released!)
>From: Ken Roberts <ken at hoverclub.net>
>To: dhcp-users <dhcp-users at isc.org>
>Cc: John Wobus <jw354 at cornell.edu>
>Date: Fri, 28 Jul 2006 14:08:51 -0500
>
>On Fri, 2006-07-28 at 13:08, John Wobus wrote:
>> 
>
>
>I hadn't known about the multiple fixed-address list until this thread,
>but one thing I think would be handy here is a wildcard like this:
>
>host joebob {hardware ethernet 00:12:3F:7C:B7:29; fixed-address
>1.2.3.4,1.2.4.4,*;}
>
>This would allow host6 to show up on any subnet (if it's a laptop for
>example) but if the server finds an address appropriate to the subnet it
>will use a fixed address.  This also lets you keep this host as a single
>address, so you would never need more than one host line per real host.

dhcpd already does exactly this and has done so for a long time. If
none of the fixed-addresses match the subnet where the request came
from, then dhcpd assigns an address out of the dynamic ranges assigned
for that subnet.

>Another interesting idea would be this:
>
>host joebob {hardware ethernet 00:12:3F:7C:B7:29; fixed-address
>1.2.*.[3-17,19];}

I think this is making the syntax too "feature rich". If you want to do
this then you can use dynamic ranges in a pool and use a class or other
way to restrict membership of that pool to a subset of clients.

>Come to think of it, I can't really see a valid reason for that second
>part unless you mapped the fourth quadrant to different firewall access
>by range and defined the authority through the host clause.
>
>However, the first example would offer a feature that I think a whole
>lot of people would want.

Like i said, already there....

regards,
-glenn