We run DB2 Community Edition DB2 v9.5.0.2, "s080811", "MI00238", Fix Pack "2". Last night, one of our databases "died", meaning, we lost connection to it. After much looking around, we finally decided to restore a backup, but the restore failed saying the backup file is corrupt. db2ckbkp lists a bunch of things missing from the backup, including LIST.OF.DATA.POOLS, history file, DATA.POOL.TABLE, Db configuration and so on. So effectively, the backup seems to be unusable. If someone can tell me how to make it usable, that'd save my life!

So anyway, the restore failed. But now, we're unable to connect to the original database anymore - the connect command gives us this SQL1119N error. I've looked through these (and other) forums and tried the things they've suggested - doing a rollforward with various options, cancelling the roll-forward, etc. No dice - the database still resolutely refuses to allow connections, giving us the same "SQL1119N A connection to or activation of database "<db alias>" cannot be made because a previous restore is incomplete or still in progress. SQLSTATE=57019" error message. How do we get this previous restore to end?

I checked the file system and there are some large files in the NODE000 sub-directories. So I am assuming our original data is still somewhere out there, on disk, and I'd love to get back to using that data. Any possibilities of that happening, do you think?

Any inputs in any of these three areas will help. Thanks!

Hi,

i don't know any to get the data out of the database when the restore started. There's a tool named "high performance unload". It's able to extract data from a Backup-Image. Don't know if it can read databases in restore- or rollforward-pending or extract data from a corrupted Backup.

It might be best to contact the IBM-Support.

Good Luck,
nvk

Unless you have some other backup image (maybe an older one), which is not corrupted, IBM support is probably your only option now.

You should have tried to restore the backup in an independent system to make sure it actually works. Once you started the backup with a corrupted image, you effectively removed the old database (so you can't use it anymore) and the corrupted image leads to the situation that the restored database can't be used either.

nvk, Knut, thank you for the responses.
nvk, I'll look for tools that like. Will post back here if I find something that works.
Knut, is there a way to fix the backup file, then? Since db2ckbkp tells me a list of things it finds missing, is there a way to introduce those things (say) into the file, so it can be restored?

Call IBM support.

See if you can unload any data using db2dart with /ddel option. If not, follow Knut's advice.