home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cis.ohio-state.edu!news.sei.cmu.edu!drycas.club.cc.cmu.edu!cantaloupe.srv.cs.cmu.edu!crabapple.srv.cs.cmu.edu!andrew.cmu.edu!<UNAUTHENTICATED>+
- Newsgroups: comp.unix.aix
- Subject: re: HELP! How to Restore Lost Logical Vol
- Message-ID: <8eeKNqb0Bwxa0IDmhY@transarc.com>
- From: Barry_Wolman@transarc.com
- Date: Sat, 5 Sep 1992 22:01:26 -0400
- References: <web30vD0Bwx3I0mEQy@transarc.com>
- Organization: Carnegie Mellon, Pittsburgh, PA
- Lines: 44
-
- This is a follow up to my posting of August 26 where I asked for help
- in restoring a "lost" 300MB file systems. Fortunately, a very
- knowledgeable and helpful IBM Field Service Quality Support analyst
- saw my posting and responded by email. With help from her, I was able
- to recover all of the file system.
-
- The recovery process was complicated by the fact that repeating the LVM
- commands that had originally been used to create the logical volumes
- in vg02 caused the logical volumes to be created at different offsets
- on the disk. "lslv -m" described the LVs as occupying the same PPs on
- the disk, but the offsets were different. My current hypothesis is
- that when vg02 was first created by our system people, they specified
- a smaller maximum number of disks for the VG than the default of 32
- that you get when using smit to create a VG; this caused the new VG
- to have a larger VGDA, which made the LVs in the recreated vg02 start
- at higher disk offsets. If I had used the same max number of disks
- for the recreated VG, my strategy of reloading the tapes and then
- recreating the Lvs probably would have worked.
-
- I finally used the following strategy to recover the file system:
-
- * recreate LVs in vg02 on disks 5 and 9
-
- * define an empty file system on lv01
-
- * reload physical backup tapes from 5 and 9 onto disks 4 and 8
-
- * copy appropriate physical 4K blocks from hdisk4 and hdisk8 into
- lv01. Knowing where the old LV was reloaded on hdisk4/8 (I had
- to write a simple program to search a disk for a LV header) and
- the allocation details of the new lv01, it was straightforward
- (but non-trivial!) to work out the series of dd commands to do the
- copy.
-
- After refilling the new lv01 with the old contents, I was able to
- mount /dev/lv01 read only and copy the files into another large
- file system.
-
- I still don't know why the volume group "disappeared". At least
- I've got the files back.
-
- Barry
-
-
-