Paula, I don't have the formula handy for calculating undo retention but a google should provide it. Have you set your retention period to large and retaining days/hours of undo info? The default is quite large. Does the batch process perform a lot of uodates that require a lot of old/new info to be retained in case of rollback? The first time i set up the undo area I accepted the defaults and consumed 6 Gig on a long running update. Check you parameters in the initxx.ora file. You can also create a new undo tablespace and assign it in the initxx.ora. Then offline drop the old one and you can then delete it from the os. Ron >>> <Paula_Stankus@xxxxxxxxxxxxxxx> 12/13/2004 1:15:44 PM >>> Have 3Gb allocated for undo retention, a few batch processes and tablespace is showing 100% full. Is this normal behavior? I am used to old rollback segments where it would be and likely revolve around poorly written batch processes. -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l