This is the current state of the database (db2 get db cfg for tmd)

Database is consistent                                  = YES
Rollforward pending                                     = DATABASE
Restore pending                                         = YES

This is the command I used to restore the database:

DB2 RESTORE DB TMD USE TSM TAKEN AT 20021103224348 REPLACE EXISTING

This finished with a successfull completion code and told me I now needed to
rollforward. So I issued:

DB2 ROLLFORWARD DB TMD TO END OF LOGS AND STOP

This then gave me a status of:

DB2 ROLLFORWARD DB TMD QUERY STATUS


                                 Rollforward Status

 Input database alias                   = tmd
 Number of nodes have returned status   = 1

 Node number                            = 0
 Rollforward status                     = DB  working
 Next log file to be read               = S0000044.LOG
 Log files processed                    =  -
 Last committed transaction             = 2002-11-02-03.44.29.000000


and that was 3-days ago. It hasn't moved. The log files are still on disk
(had an error in the db2uext2.c). The environment variables are set
correctly:

DSMI_CONFIG=/db2/db2inst3/tsm/dsm.opt
DSMI_DIR=/usr/tivoli/tsm/client/api/bin
DSMI_LOG=/db2/db2inst3/tsm

I can see in my server activity log the client connections, but the TAKEN AT
backup I know is on tape. It never asked for a tape mount! But it gave me a
successfully completion of the RESTORE.

The backups of the other instances are working just fine, and until I hosed
this one up with the restore, the backups were completing nightly.

I'm at a point now where I'm thinking of DB2STOP'ing the instance and
restoring the whole /db2/db2inst3/db2inst3 directory back to a point-in-time
from before I FUBAR'd it.


Bill Boyer
DSS, Inc.