Thanks Stefan, I appreciate the feedback.
I will run another test with TEMP_UNDO_ENABLED set to TRUE.
-----Original Message-----
From: Stefan Koehler [mailto:contact@xxxxxxxx] ;
Sent: Monday, June 05, 2017 12:43 PM
To: Hameed, Amir <Amir.Hameed@xxxxxxxxx>; 'ORACLE-L' <oracle-l@xxxxxxxxxxxxx>
Subject: Re: INSERT into temp table causing "log file switch (checkpoint
incomplete)"
Hey Amir,
1) Because of UNDO. However you can change this behavior with 12c for GTTs
(parameter TEMP_UNDO_ENABLED -
http://docs.oracle.com/database/121/ADMIN/undo.htm#ADMIN13741).
2) It depends :-)
Best Regards
Stefan Koehler
Independent Oracle performance consultant and researcher
Website: http://www.soocs.de
Twitter: @OracleSK
Upcoming online seminar: http://tinyurl.com/17-06-13-Shared-Pool-Internals
"Hameed, Amir" <Amir.Hameed@xxxxxxxxx> hat am 5. Juni 2017 um 18:27
geschrieben:
Hi,
What I am trying to understand is:
1. Why would insert into a global temporary table generate so much
redo?
2. Initially there were four redo log groups but I added another four
to see if that would alleviate the “log file switch (checkpoint
incomplete)” wait but it did not help. Is that because even four DBWR
processes are not enough to perform checkpoints in a timely manner?
Thanks
Amir
