Case 1: if system panicked (like the panics I sent ksymoops'd oopses of), Alt+SysRq+S will sync all discs until my 00:0a (/dev/sda10, /opt/data) and then send another panic (I also sent ks
he xfs flush and flush invalidate calls down the what the rest of the kernel uses. The following file(s) were checked into: bonnie.engr.sgi.com:/isms/slinx/2.5.x-xfs Modid: 2.5.x-xfs:slinx:
ixup, pending patch from Andrew Morton, removes the case where we tossed cached data but never flushed it. This has been submitted to Linus already. The following file(s) were checked into:
I have an xfs filesystem on an lvm volume on an i386 ide disk. xfsdump seems to happily work: date=`date +%y-%m-%d` xfsdump -l 0 -p 30 -L home.full.$date -M home.full.$date - /home I wanted to save t
Hi Matthew, Yes. I can't see why not. I just tried it and it worked for me :) I would presume the "Bus error" is just the program crashing out from a SIGBUS signal with some address access problem su
I created a small volume for testing and ran the simplest dump/restore I could think of: $ mkfs -t xfs -f /dev/homerdata/testvol $ mount /dev/homerdata/testvol /mnt/xfstest/ $ touch /mnt/xfstest/test
I had 2.1.3 (Debian Potato) and now use 2.2.5 (Debian Woody). I assume it was the libc upgrade that did it. A simple recompile of xfsrestore solved the symptoms. -- Matt Rose |Necessity is the plea f