We help IT Professionals succeed at work.

Unable to kill Incomplete Db2 Restore: How can I use extreme force/Abort to release the databse?

Enyimba asked
Hi Database Gurus,

I did a redirected restore on a DB2 database using a backup image from a production instance, generated the script, made neccessary changes to the script - ran the set tablespaces commands, changed the container paths and all looked well. After runnig the redirect script, I observed that it appeared to have hanged. Now I am in the situation that if I issue a restore complete it says "DB21080E No previous RESTORE DATABASE command with REDIRECT...." and if I try to do anything else that involves a connection (including trying to drop the damn database) I get "SQL1119N A connection to or activation of database "RAPMT02" cannot be made because a previous restore is incomplete. SQLSTATE=57019".

I have tried "DB2 restore...ABORT..." with no success. I do a ps -ef | grep - restore, get the PID and have also tried to kill
-9 PID but that did not allow me to connect to the database either giving me the "DB21080E No previous RESTORE DATABASE command with REDIRECT....".

I will like not to have to drop the Database. So anyone have any ideas? Can I salvage the restore? If not how can I remove the damaged db so I can start again? If possible, I appreciate an answer tonight as I am working to fix this problem on a deadline.

Thanks all
Watch Question

Data Warehouse / Database Architect
Hi Enyimba,

You need to abort the previous restore.

db2 "restore database {alias} ABORT"

On older versions of DB2 you sometimes had to include some of the load parameters.  But try it like the example to start.

Good Luck!


thanks kdo