At 11:11 18-2-2004 -0800, Joshua Schmidlkofer wrote:
I see that running xfs_db is dangerous. I have a system that I ran
xfs_repair (-n -f) on a live filesystem, and it was a Bad Thing(tm). Is
that the case? not a lot was going on, but I got some really nice call
traces, =(. Then 990's on directory access, messages about corrupted
inodes. I initiated a reboot, but not before saving the logs
Fixing a filesystem.
1. Reboot the server with your favorite rescue disk. Mine is the lnx-bbc disk.
2. Mount the filesystem, unmount the filesystem (so the log is flushed).
3. run xfs_repair -n to see what it wants to do. If it looks normalish you
can run xfs_repair without -n. If xfs_repair recovers file and puts them in
lost+found they will be unlinked and found again if you run xfs_repair
4 It should be fixed. Reboot the server. Files might end up in lost+found
if there are big problems.
I am trying to get data off it on to the stand-by, but Am I Borked?
I don't think everything is hosed. But try above suggestion. Lnx-bbc disk
has network support and other neccesary utilities to get data off.
I don't make sense, I don't pretend to either. Questions?