On Jan 5, 2012, at 1:40 PM, John A. Sullivan III wrote:
On Thu, 2012-01-05 at 12:57 -0500, Rick Thomas wrote:A limited amount of redundancy is good. If one goes down, the networkThe problem is when they are administering addresses in the same range. I've not configured DHCP for a long time so maybe this is common now andcan still limp along. Anyway, that's the theory. Rick On Jan 5, 2012, at 10:07 AM, Osamu Aoki wrote:I'm not the OP, but I do have this problem. When I try to do aninstall (wheezy) on a network with two DHCP servers, the installer'sdhcp-client never seems to get an IP address -- even though the two servers are both responding and both giving the same IP address.Then why have 2?the problems have all been resolved but, in the past, if one wantedredundancy, one would administer different ranges on the same subnet sothat there would be no conflicts. Hope that helps - John
Yes, that does seem to fit.The two servers have different ranges for their "dynamic" clients (i.e. transient laptops with no fixed IP address) but for the "static" clients (desktops with a fixed IP address) there is only one address, so both servers have to provide the same address.
So... when I do an install for a machine with an unknown Ethernet Mac address -- hence getting it's IP from one or another of the "dynamic" ranges -- all goes well. It's only when I attempt to install a machine with a known Mac address (hence a single static IP) that I run into trouble.
Somehow, I would have thought it should be the other way round -- conflicting responses would cause problems, not two responses that both say the same thing. Sigh! Just shows how much I know...
So is there a way to have both redundancy *and* reliable installs? Thanks for any help, RickPS: As others have noted, it seems to affect only the installer's DHCP client. After the reboot, there's no problem -- with either static or dynamic IP...