<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffcc" text="#000000">
<br>
<br>
On 03/06/10 10:57, Simon Hobson wrote:
<blockquote cite="mid:p06240814c82d1d38792a@simon.thehobsons.co.uk"
type="cite">Marc Chamberlin wrote:
<br>
<br>
<blockquote type="cite">
<blockquote type="cite">5.) I have been doing this a number of
years. I simply assigned both interfaces the same IP, with the effect
that as they use only one interface at a time, everything worked even
when switching interfaces during a download.
<br>
<br>
Whether this is in accordance with the RFC, I am not sure. It worked
for me and was simple.
<br>
<br>
</blockquote>
Thanks Sten for your comments.. I would be interested in learning more
about your approach, but keep in mind that I am trying to accomplish
this task in the context of assignment of static IP addresses based not
only on the hardware interface addresses of client machines, but on the
operating system that is running on the client system also...
<br>
</blockquote>
<br>
That's the problem. I assume Sten is doing the same as I have for my
laptop :
<br>
<br>
host "laptop-wired" {
<br>
hardware ethernet aa:bb:cc:dd:ee:ff ;
<br>
fixed address 192.168.1.n ;
<br>
}
<br>
host "laptop-wireless" {
<br>
hardware ethernet uu:vv:ww:xx:yy:zz ;
<br>
fixed address 192.168.1.n ;
<br>
}
<br>
</blockquote>
This is indeed what I have done, the effect is that even if two
interfaces are configured to use the same IP, only one of them will be
active at any one time. The only level that may be confused is ARP, but
that solves itself fast and traffic continues.<br>
<br>
My point is that if you give one address to more than one interface
according to your rules, that will not likely confuse anything. This
should help remove one level of complication from your problem; but of
course not solve it completely.<br>
<br>
Basically use OS-info to determine the IP, then assign it to whatever
interface(s) is there, you don't have to give separate addresses to
each interface. How this exactly fits into the final solution I don't
know.<br>
<blockquote cite="mid:p06240814c82d1d38792a@simon.thehobsons.co.uk"
type="cite"><br>
This seems to work for my Mac and it's never complained when I forget
to turn off the wireless at home. But as you have figured out, it won't
work in a situation using classes etc since the server would tell you
"no free leases". This is the same problem faced by people (such as
cable internet providers) doing address assignment based on Option 82 -
if the end user changes their equipment, then the MAC address no longer
matches and there will be no lease free in that single address pool
until the old lease expires (or is manually ended)
<br>
<br>
As an aside, IIRC you commented that this error message was misleading
in an earlier message. I guess you are partially right, but also the
error message is correct if you think of it as meaning "no free leases
**given the constraints applied**".
<br>
<br>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Best regards
Sten Carlsen
No improvements come from shouting:
"MALE BOVINE MANURE!!!"
</pre>
</body>
</html>