I have a deadline to meet. I must migrate all my servers, and clients to a private IP space. Servers must be done — if at all possible — by September 30th. Clients by October 30th. That is, this year. Problem is, I have never done anything close to it, and there are so many things that need to be done at once, I don’t know where to start.
I have a Linux DNS/DHCP to change (Internet facing one). I have an AD DNS to change (AD, as AD integrated zones. Three servers, actually. All for internal AD use). I have 20 servers to change (file, print, DFS, web multi-homed, etc). There are close to 1000 clients as well, but those, once DHCP is migrated, will just fall in place. But… I also have special areas, like laboratories, for example, that are currently assigned to a specific IP segment (using their MAC).
Thinking where to start.
I tried adding a new IP to the Linux box (the one with DNS/DHCP), to start at a point. It is breaking DNS when I do that. DNS resolution stops as soon as “
ifup ifcfg-eth0:0” is issued. I am drawing a blank here, since
ifcfg-eth0:0 is simply a copy of
ifcfg-eth0, reflecting the other IP, and the
DEVICE=eth0:0 instead of
DEVICE=eth0. All by the book. How is that causing DNS to malfunction?
I will come back and update, or follow up on a different post. Open to suggestions though.
Changing the name to
ifcfg-eth0:1 worked. Was
ifcfg-eth0:0 getting confused with eth1 (the second — and disabled — NIC)? I am not sure, but certainly the naming had something to do with it. I never did this on servers with two NIC, and most of the HOWTOs do not take that into consideration.