Database Restore after hardware failure

Vitalogy301

Active Newcomer
Joined
Dec 18, 2012
Messages
7
Reaction score
0
Points
0
I have an old TSM server running 5.4.
One of my database drives has failed (replacement on the way).
I cannot start dsmserv.exe.
See error messages below. Obviously Drive F is my failed drive.
I found Tech Note 1268764 which references this error.
It suggests:
Resolution:

+ Create a new directory on any available drive on the machine.

+ Share that directory

+ Grant "Full Control" to Everyone in the Share permissions

+ Allow "Full Control" to the Admin User in the NTFS Security

+ Map the new share to the original drive letter, and create the necessary directories

+ Copy the dislocated db volume to the new location

+ Restart the TSM server.

Am I headed in the right direction?

Thanks!


ANR9969E Unable to open volume F:\TSMDB\DB1.DSM. The most likely reason is that
another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\DB2.DSM. The most likely reason is that
another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\D4050454.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\C5050454.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\D3880984.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\C3880984.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\D4092325.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\C4092325.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\D8657565.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR9969E Unable to open volume F:\TSMDB\C8657565.DBV. The most likely reason is
that another TSM server is running and has the volume allocated.
ANR0259E Unable to read complete restart/checkpoint information from any
database or recovery log volume.
 
Hi,

question are - do you have those volumes/files? Is it complete set of the DB/RLOG volumes needed? Was you DB/RLOG mirrored? Do you have recent DB backup?

Harry
 
Hi Harry,
Yes, the DB/RLOG was mirrored - Drive F and Drive G
I have a recent DB backup
Here's what's on G:\
tsmdbm
\C7038666.DBV
\db1.dsm
\db2.dsm
TSMLOG
\LOG1.DSM
 
Hi,

from here it seems you used to have mirrored DB - those .DBV files seem to be created by automatic spacetrigger - and it seems they were not created in both locations.
How does the dsmserv.dsk file look like? What about MIRRORREAD and MIRRORWRITE parameters in the dsmserv.opt? Are they defined?

Harry
 
#dsk_comment#page_shadow_token:1120810165051
F:\TSMDB\DB1.DSM
F:\TSMDB\DB2.DSM
G:\TSMDBM\DB2.DSM
G:\TSMDBM\DB1.DSM
F:\TSMDB\D4050454.DBV
F:\TSMDB\C5050454.DBV
F:\TSMDB\D3880984.DBV
F:\TSMDB\C3880984.DBV
F:\TSMDB\D4092325.DBV
F:\TSMDB\C4092325.DBV
F:\TSMDB\D8657565.DBV
F:\TSMDB\C8657565.DBV
G:\TSMDB2D7038666.DBV
G:\TSMDBM\C7038666.DBV
G:\TSMLOG\LOG1.DSM
 
Hi,

confirmed - the DB is lost - in case the DB was mirrored you should have the same number of volumes on the both disks - you have not (F: 10 files, G: 5 files)
Time to restore the DB.

Harry
 
MIRRORREAD
*
* ********************************************************************
* MIRRORRead
*
* Specifies the mode used for reading recovery log pages or data base
* log pages
*
* Syntax
* +------------------+-----------+----------------------------------+
* | MIRRORRead | LOG | DB | Normal | Verify |
* +------------------+-----------+----------------------------------+
*
MIRRORRead DB Normal
*
* =====================================================================
*
* MIRRORWRITE
*
* ********************************************************************
* MIRRORWrite
*
* Specifies how mirrored volumes are accessed when the server writes
* pages to the recovery log or database during normal processing.
*
* Syntax
* +------------------+-----------+----------------------------------+
* | MIRRORWrite | LOG | DB | Sequential | Parallel |
* +------------------+-----------+----------------------------------+
*
MIRRORWrite DB Sequential
*
 
Do I have to wait until I get the replacement drive or can something be done now? Can I restore the database if I'm not able to run dsmserv?
Oh...and thanks for your assistance!
 
Hi,

if you have enough free space on the G: drive, then you can restore the DB now.
Follow the guidelines in the Admin guide
http://publib.boulder.ibm.com/infocenter/tivihelp/v1r1/topic/com.ibm.itsmcw.doc_5.4/anrwgd54.pdf
and the reference:
http://publib.boulder.ibm.com/infocenter/tivihelp/v1r1/topic/com.ibm.itsmcw.doc_5.4/anrwrf54.pdf
(look for Appendix A - TSM Utilities)

What you have to do is to create additional DB/Rlog volumes to accomodate the DB to be restored, then format it and restore the DB. All the commands are dsmserv<command> <parameters> - see the reference for exact syntax.

Harry
 
Thanks Harry. I'm still confused about not being able to run dsmserv.exe but I'll read thru the information and give it a try.
 
Hi,

when running dsmserv.exe without parameters, TSM server is started (on the foreground).
However dsmserv.exe can be run with various maintenance options - these are usually used for recovering/dumping/unloading/loading/formatting the database.
You need "dsmserv format" and "dsmserv restore db"
BEWARE - save your original volhist.out and devconfig.txt files.

Harry
 
Through some trial and error the database is restored and the server is up and running again. Yea!
I am seeing ANR1331E errors on migration.
Ran AUDIT VOLUME FIX=YES
I'm now seeing many ANR1168W Migration skipping damaged file on volume...

Any suggestions? Thanks!
 
Hello,

seems the files on the volume(s) are damaged but there is an existing copy in one (or more) copystgpools. Use "move data" command to get all readable data off the media and then issue "restore volume" to recover the primary copies from the copy stgpools.

Harry
 
Back
Top