I can't reproduce it as reliably, but it's the same stack trace as Bugzilla id #239520.
Maarten Broekman -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jarod Wilson Sent: Wednesday, August 29, 2007 8:54 PM To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list Subject: Re: [rhelv5-list] Help needed with kdump on cciss devices Broekman, Maarten wrote: > Thanks for the info. Does this also mean that kdump isn't able to read > /proc/vmcore off of them (cciss partitions) when it's trying to dump? Reading /proc/vmcore has nothing to do with any partitions, it is a psuedo-block-device interface to memory, it doesn't reside on any partition. > I > ask because I changed from writing to a filesystem to scp'ing to a > remote system as a trusted user and dumping there. In that situation, whether or not any partitions are mounted doesn't matter. You read in from /proc/vmcore and write to a remote system, nothing ever touches your local file system. > Since I made that > change I've only managed to get "Kernel panic - not syncing: Fatal > exception" errors after a stack trace which includes :cciss: calls > (:cciss:cciss_getgeometry, :cciss:cciss_init_one, and :cciss:sendcmd). Hm... That's a new one to me. If its something you're able to reproduce, please do file a bug. > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Jarod Wilson > Sent: Wednesday, August 29, 2007 2:06 PM > To: Red Hat Enterprise Linux 5 (Tikanga) discussion mailing-list > Subject: Re: [rhelv5-list] Help needed with kdump on cciss devices > > Broekman, Maarten wrote: >> I'm currently trying to get RHEL5 kdump to work with my cciss-based >> internal drives. Every time I point kdump at my /var/crash partition >> (ext3 /dev/cciss/c0d0p5), kdump complains that the filesystem isn't >> ext2. Anyone had luck getting this to work? > > Both cciss and cpqarray controllers do things a little different than > the norm w/their partitions, which throws the rhel5 kdump for a bit of a > loop when your try to specify a specific partition. We've got this fixed > properly for rhel5.1. > -- Jarod Wilson [EMAIL PROTECTED] _______________________________________________ rhelv5-list mailing list [email protected] https://www.redhat.com/mailman/listinfo/rhelv5-list _______________________________________________ rhelv5-list mailing list [email protected] https://www.redhat.com/mailman/listinfo/rhelv5-list
