[Techtalk] Dump constipated and kernel panic (solved)

[Conor Daly]

Well, solved sort of...

I did an fsck on all filesystems after the latest crash and dump went from
crashing the machine to segfaulting instead.  I have it figured down to
the "-z9" (max compression) argument, dump runs happily with "-z2".

Conor

On Tue, Jul 16, 2002 at 10:37:48AM +0100 or so it is rumoured hereabouts, 
Conor Daly thought:
> Sorry about the subject line but I couldn't resist :-)
> 
> Running RH7.2 kernel 2.4.7-10 ext3 filesystems / on /dev/hda2, 
> /home on /dev/hda4 with dump v0.4b29 (upgraded from 0.4b22) and getting 
> a kernel panic
> 
> command is
> 
> dump -0 -B 665600 -f /home/dump.dat -F "dump2CDR" -z9 /
> 
> where /home is a seperate filesystem and dump2CDR burns /home/dump.dat to
> CDRW via mkisofs.  The dump proceeds to do the first section, builds
> /home/dump.dat and hangs.  Sometimes the system hangs hard, bouncing the 
> power is the only option, sometimes I get this kernel oops.  I thought 
> initially it was related to the CDRW drive (NEC something) but same 
> symptoms appear with a different CDRW drive (LG).  
> I upgraded dump from 0.4b22 to 0.4b29 (latest on dump.sourceforge.net)
> with the same result.  After a hard hang, on reboot, /dev/hda2 fscks and 
> returns OK while I haven't done a fsck on /dev/hda4 (init does a "recovering 
> journal" instead).  
> 
> I seem to remember Linus said something about not using dump with ext3
> filesystems and I wonder if it's related.  Haven't tried a kernel upgrade yet.  
-- 
Conor Daly <conor.daly at oceanfree.net>

Domestic Sysadmin :-)
---------------------
Faenor.cod.ie
  4:22am  up 55 days, 13:40,  0 users,  load average: 0.00, 0.00, 0.00
Hobbiton.cod.ie
  4:11am  up 29 days,  3:33,  1 user,  load average: 0.11, 0.48, 0.31