1) Make sure you save a copy of your current volume history file to some other
location before restarting ADSM for the first time. ADSM will rewrite it when
it starts up. If you restored your ADSM database to some point in time (i.e.
you did not run AUDITDB against the current database), then the volume history
file will be rewritten with older data and the newer entries from after the
time you took the database backup will be lost. You will need those newer
entries. So saving the volume history file to another file/location is
important.
2) Add the NOMIGRRECL option to your ADSM server options file (just add a line
that says NOMIGRRECL. There are no other parameters for this option). This will
prevent migration and reclamation from running.
3) Start the ADSM server, and immediately issue the DISABLE command to prevent
backup-archive client access. Administrators can still access the server.
4) Run AUDIT VOLUME for each disk storage pool volume in your storage pools.
NOTE: Steps 5, 5a, and 5b are not necessary if you audited a current database
(i.e. you did not restore anything), or if you ran RESTORE DB and restored the
database to its most current state (your recovery log would have had to been in
roll forward mode, and it would have had to be intact when you did the RESTORE
DB for this to occur).
5) Examine the backup copy of the volume history file you made in step 1 above.
In particular, you need to focus on entries from *after* the time your database
was restored to. In other words, if you restored your database to the state it
was in as of last Thursday at 13:00, then entries after 13:00 last Thursday are
the ones you need to pay attention to with regard to items 5a and 5b below:
5a) ... For any entries that show volumes deleted from ADSM storage pools
(STGDELETE), you need to determine whether those volumes have been written over
(either by some other application of ADSM itself). If they have been written
over, then you need to do DELETE VOLUME with DISCARDDATA=YES for each volume
that falls in this category.
5b) For any entries that show volumes added to storage pools (STGNEW), you need
to return these to your tape management system (external to ADSM) as scratch.
6) Once you've completed these items, stop the ADSM server, take out the
NOMIGRRECL option, and restart. You should be back in business.
Andy Raibeck
IBM Storage Systems Division
ADSM Client Development
e-mail: storman AT us.ibm DOT com
ADSM-L AT VM.MARIST DOT EDU
03/18/98 12:03 PM
Please respond to ADSM-L AT VM.MARIST DOT EDU @ internet
To: ADSM-L AT VM.MARIST DOT EDU @ internet
cc:
Subject: AuditDB
We have FINALLY completed the salvage commands including AUDITDB. Is there
anything else we need to do before starting the ADSM server, such as restore
the volhist file on ADSM's 191 disk?
We've waited five long days for this process to complete and we're very eager
to see if it worked.
Thanks to everyone for their help, suggestions, and encouragement. The tale
is about to be told.
|
