Corrupt Root Logical Volume
678641Mar 30 2009 — edited Mar 31 2009Hey Folks,
I'm having a bit of a problem with the Root Logical Volume / Volume Group on one of our Servers (OEL 4 - Release 7 running on Oracle VM 2.1.2). One of our SAN Administrators pulled the disk from the system by accident, and has since re-presented it to the server but now when it boots I'm getting errors on the Root Logical Volume.
It boots in to maintenance mode and I've tried fsck'ing the root file system but it gives errors about the first SuperBlock...
"Attempt to read block from filesystem resulted in short read while trying to open /dev/VolGroup00/LogVol00
Could this be a zero-length partition? "
I've tried specifying the next SuperBlock and the e2fsck runs through repairing corrupt inodes, but it doesn't make a difference, I still get errors on bootup. Any other LVM commands give an error to say that 'File Descriptor XX not closed'.
I've booted from an OS DVD in to Rescue Mode and it mounts the Volume Groups and I can still see the files are present.
Does anyone know of anything I can do? Are their commands I can run in Rescue Mode to fix corrupted file systems? I'm off to do some googling on the subject but if anyone could offer advice it would be much appreciated.
Thanks
James