RE: Rebuilding MLOG tables

2003-06-13 Thread jo_holvoet
/2003 22:41 Please respond to ORACLE-L To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Subject:RE: Rebuilding MLOG tables I don't think we can go with the truncate table thing since there is too much weirdness around here in when a client

RE: Rebuilding MLOG tables

2003-06-13 Thread Stephen Lee
Didn't know about this one. Thanky thanky. So now we have dbms_refresh, dbms_repcat, and dbms_mview (more?) each with its own bucket of procedures. It gives one the impression that there is some significant developer turnover at Oracle with each new batch of programmers imposing their own

RE: Rebuilding MLOG tables

2003-06-13 Thread Stephen Lee
Mebbe this has something to do with the implicit commit stuff associated with DDL (iirc)? I keep coming back to what Einstein said (I think it was him): Education is what is left over after you have forgotten everything you have learned. (quote might not be exactly correct) -Original

Re: Rebuilding MLOG tables

2003-06-13 Thread Arup Nanda
You are very welcome. I agree, Oracle must have had a wave of those PHBs in their development side each with their own trumpet to blow and each left his or her legacy with a new package. Otherwise why they cose to have so many of these packages to do a few simple, very correlated things beats

RE: Rebuilding MLOG tables

2003-06-13 Thread Stephen Lee
recipients of list ORACLE-L Subject: Re: Rebuilding MLOG tables You are very welcome. I agree, Oracle must have had a wave of those PHBs in their development side each with their own trumpet to blow and each left his or her legacy with a new package. Otherwise why they cose to have so

Re: Rebuilding MLOG tables

2003-06-12 Thread Arup Nanda
Stephen, I do it all the time. Actually you don't have to lock the table; you may simply quiesce the table, meaning no transations will be allowed. Steps: Quiesce the table Apply all the pending logs in the deferred trans queue on secondary database Truncate The MLOG$ table. No issues; in fact

RE: Rebuilding MLOG tables

2003-06-12 Thread Sarnowski, Chris
I don't think it will do quite what you want it to, since the 'alter table move' statements are DDL so will release the lock. I just tried this experiment in 2 SQLPlus windows: SQL 1.1 lock table blah in exclusive mode; returns with Table(s) Locked. SQL 2.1 insert into blah values (1); (this

RE: Rebuilding MLOG tables

2003-06-12 Thread Stephen Lee
I don't think we can go with the truncate table thing since there is too much weirdness around here in when a client -- and there are multiple clients -- might update: Network problems, box crashed, sunspots (Don't forget about the sunspots!). So if there are entries still hanging around in the

RE: Rebuilding MLOG tables

2003-06-12 Thread Stephen Lee
:[EMAIL PROTECTED] Sent: Thursday, June 12, 2003 3:38 PM To: Multiple recipients of list ORACLE-L Subject: RE: Rebuilding MLOG tables I don't think it will do quite what you want it to, since the 'alter table move' statements are DDL so will release the lock. I just tried this experiment

Re: Rebuilding MLOG tables

2003-06-12 Thread Arup Nanda
To: Multiple recipients of list ORACLE-L Subject: RE: Rebuilding MLOG tables I don't think it will do quite what you want it to, since the 'alter table move' statements are DDL so will release the lock. I just tried this experiment in 2 SQLPlus windows: SQL 1.1 lock table