On 2011-05-04, James Chase <ja...@chasecomputers.net> wrote: > Hi, > > I recently upgraded to 4.9 and everything seemed OK between my master > and backup firewalls setup with CARP and state sharing. The firewalls > were setup strange, the backup had the internal IP of .25 and the master > had .26. I wanted to swap this so that the master firewall IP would be > in numerical order, so I changed the hostname.x file and rebooted them > both at the same time. > > After that I had continuous issues with the backup firewall. It just > kept restarting, so I switched the IP's back. The problem continued, and > I found that it would only boot without crashing if I let it boot with > the network cables unplugged and then plugged them in after it was up. > However, trying to connect to that firewall with SSH causes the system > to crash immediately (it often makes a click sound right when you do it > and then hard reboots). It seemed like it functioned when I tested the > carp failover though! > > I brought the firewall home and it works fine away from that network. > The crash dumps don't appear to be understandable by the everyday man.
Without the information that's in them, nobody is going to be able to even guess what might be wrong. Particularly important:- - any panic or fault message, - output from 'trace' and 'ps' commands, probably also 'sh reg' - at what point does it crash? what is displayed on screen before any panic/fault message? > Is there some sort of disk caching or other area that could cause the IP > change to have created a problem? I only changed the hostname file for > the interfaces I wanted to swap (internal interface and the pfsync > interface) so I'm not sure where the problem would be. It's not disk caching.
