Re: Composite partitioned table purge causes need for index rebuild
From: Matt <mccmx_at_hotmail.com>
Date: Wed, 16 Sep 2009 15:40:02 -0700 (PDT)
Message-ID: <889d2f17-453d-42b6-ae34-bfee1bd3087d_at_x5g2000prf.googlegroups.com>
The rebuild successfully survived various failures (kill -9, tablespace full etc) but the kill -9 failure required a manual clean up of the data dictionary before another online rebuild could be attempted. This is because the journal table remains after the crash and the next rebuild attempt tries to create a journal table which already exists.
Date: Wed, 16 Sep 2009 15:40:02 -0700 (PDT)
Message-ID: <889d2f17-453d-42b6-ae34-bfee1bd3087d_at_x5g2000prf.googlegroups.com>
The rebuild successfully survived various failures (kill -9, tablespace full etc) but the kill -9 failure required a manual clean up of the data dictionary before another online rebuild could be attempted. This is because the journal table remains after the crash and the next rebuild attempt tries to create a journal table which already exists.
This can be cleaned up by running:
begin
dbms_repair.ONLINE_INDEX_CLEAN(<object_id>); end;
thanks again for your response. Received on Wed Sep 16 2009 - 17:40:02 CDT