[Veritas-bu] Consolidation Interupt and Failure
2002-07-16 18:29:09
Hi
We have a HSM system that migrates two copies: one to Optical Platter
and one to DLT tape. On Monday we had a system crash. Unfortunately I?d
just started the consolidation of an optical platter.
Once we had the system back up I ran
migrc ?L
Then started the HSM daemons. Functionality seemed to be okay. Because
the Optical Platter consolidation had only just started I thought that
all I needed to do was remove the CONS_INVOL and CONS_OUTVOL files from
the database directory; then rerun the consolidation of the same
platter?..wrong !
The consolidation ran, but finished like this:
00077A gran 52: fh=3e8Mffb06 off=x0
00077A gran 53: fh=3e8Mffb08 off=x0
00077A gran 54: fh=3e8Mffb0a off=x0
00077A gran 55: fh=3e8Mffb0c off=x0
00077A gran 56: fh=3e8Mffb16 off=x0
00077A gran 57: fh=3e8Mffb1a off=x0
CONSOLIDATION SUCCESSFUL ***
**** merging FHDB database under progress
mv: /opt/openv/hsm/ssa_1/workdir/ssa_1.DVDB.ed3 not found
**** Input media cleaning under progress
**** FHDB database cleaning under progress
**** VOLDB database cleaning under progress
FREE to RE-REGISTER volume ===> 00007A op
00007B op
=================================================
NEW volume Generated ===> 00136B op
00077A op
Finished
Although it successfully moved the files physically to the target
platter it did not merge the new file handles into FHDB. So all the
files (about a hundred) that I?d tried to consolidate do not have system
handles to optical platter. This is not actually that serious in itself
as I still have DLT file handles. I can use HSM to bring them back into
cauche then remigrate them.
The challenge is getting my next consolidation to work ! I did a check
of the workdir and found the following files:
-rwxr-xr-x 1 root other 0 Jul 15 11:38
CONS.DVDB.FHDB.6ced
-rwxr-xr-x 1 root other 45076 Jul 16 15:20 CONS.DVDB.FHDB.ed3
-rw------- 1 root other 6 Jul 16 15:37 mig.lck
-rw-r--r-- 1 root other 0 Jul 15 19:01
ssa_1.copydb.ad.1.library
-rw-r--r-- 1 root other 0 Jul 15 19:01
ssa_1.copydb.dt.1.library
-rw-r--r-- 1 root other 0 Jul 15 19:01
ssa_1.copydb.op.1.library
-rw-rw-r-- 1 root root 0 Jul 15 19:01
ssa_1.copydb.op.1.library.temp.19072
-rw-r--r-- 1 root other 0 Jul 15 19:01
ssa_1.copydb.op.2.library
-rw-r--r-- 1 root other 0 Aug 28 2001
ssa_1.dvdb.sort.1110
-rw------- 1 root other 6 Jul 16 15:37
ssa_1.migcons.running
-rw------- 1 root other 6 Jul 15 19:01
ssa_1.migrc.running
-rw-r--r-- 1 root other 0 Jan 4 2002
ssa_1.migsweepm.11608
-rw-r--r-- 1 root other 0 Oct 5 1999
ssa_1.migsweepm.13035
-rw-r--r-- 1 root other 2 Jul 15 19:01 ssa_1.p_badness
-rw-r--r-- 1 root other 3 Jul 15 19:01 ssa_1.s_badness
I would hazard a guess that by removing the CONS* files up the top and
getting rid of ssa_1.copydb.op.1.library.temp.19072 a subsequent
consolidation would work. This is open heart surgery I admit. Has
anybody battled a similar issue ?
Terry
|
<Prev in Thread] |
Current Thread |
[Next in Thread> |
- [Veritas-bu] Consolidation Interupt and Failure,
Terry Newnham <=
|
|
|