Howdy,
We are experiencing a bizarre problem under Linux attempting to restore a DB2 database. The backup was created on this machine, and is being restored to the same machine, but changing the name of the database, and redirecting the tablespaces. The system is running DB2 Express C V9.5 and RHEL5 with latest updates
The restore commands we are trying to run are as follows:
-
RESTORE DATABASE CARDS FROM "/opt/rccs/data/backups" TAKEN AT 20081105140449 TO "/opt/rccs/data" INTO bkl2 NEWLOGPATH "/opt/rccs/data/bkl2/dblogs" WITH 2 BUFFERS BUFFER 1024 REDIRECT PARALLELISM 1 WITHOUT PROMPTING;
-
SET TABLESPACE CONTAINERS FOR 7 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/cards");
-
SET TABLESPACE CONTAINERS FOR 12 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/changedata");
-
SET TABLESPACE CONTAINERS FOR 5 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/config");
-
SET TABLESPACE CONTAINERS FOR 8 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/transactions");
-
SET TABLESPACE CONTAINERS FOR 4 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/user");
-
SET TABLESPACE CONTAINERS FOR 9 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (PATH "/opt/rccs/data/bkl2/ts/tranmessage");
-
SET TABLESPACE CONTAINERS FOR 11 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (FILE "/opt/rccs/data/bkl2/ts/TSASNAA" 124160);
-
SET TABLESPACE CONTAINERS FOR 10 REPLAY ROLLFORWARD CONTAINER OPERATIONS USING (FILE "/opt/rccs/data/bkl2/ts/TSASNUOW" 57344);
-
RESTORE DATABASE CARDS CONTINUE;
-
ROLLFORWARD DATABASE bkl2 TO END OF LOGS AND COMPLETE;
-
This script, executed from a shell via a "db2 -tf" command, works perfectly on another, similarly configured machine when it was used to restore an existing database to a new name. When the same script was executed on the problem machine the first command in the script, the RESTORE command, fails with an SQL1032N error. This error is "No start database manager command was issued", yet if we execute a db2start command, we get the message "The database manager is already active".
We are able to create (and drop) databases from the same shell.
The output from the db2level command is:
DB21085I Instance "dbadmin" uses "32" bits and DB2 code release SQL09050" with level identifier "03010107".
Informational tokens are "DB2 v9.5.0.0", "s071001", "LINUXIA3295", and Fix Pack "0".
Product is installed at "/opt/IBM/db2/V9.5".
and is identical on both machines.
I'm out of ideas as to a solution. Any help will be appreciated.
Mike Laris
mlaris@sumx.com
