Hi > I am curious - why are you forced to keep > segments in specifically named tablespaces.
This Concorde XAL application has an internal repository which stores information about physical segment structures (as tablespace name, initial & next extent etc). And it can start reorganizing segments without any warning if it finds some inconsistencies between data dictionary and its own information. So, we decided itīd be easier to move data back to original tablespaces instead of starting re-configuring the app (we have whole weekend to do the job). > I think the ideal is to MINIMIZE records_per_block on > such tables, then set pctfree to zero, then MOVE them, > then set pctfree back to what it was. But if you do have > any such tables, you might want to experiment. Good point about pctfree, but how could minimize records_per_block help in this case? Iīve used it for optimizing bitmap indexes, but for table moving? Tanel.
