RE: Big Update/DML
Date: Wed, 26 Aug 2020 14:30:31 +0000
Message-ID: <643f6e3a7172480c81f6bc0839badf7b_at_imcnam.ssmb.com>
Be careful with how you do parallelism. Done correctly it will speed things up. Done incorrectly and you will have a locking nightmare. Are you updating the columns with the same value? If so, the default value option might be very useful.
Liz
From: [freelists.org] oracle-l-bounce_at_freelists.org <oracle-l-bounce_at_freelists.org> On Behalf Of [freelists.org] Sanjay Mishra
Sent: Tuesday, August 25, 2020 11:29 PM
To: oracle-l_at_freelists.org
Hi Everyone
I am working on 19c and need to do one time update on multiple tables containing 3-4 Billions records and some tables are Compressed for OLTP and some are uncompressed. Tables have multiple columns but updating only one new column added with data from another column from the same table. Environment is on Exadata with Buffer Cache of 60G and CPU_count of 30
Update using high Parallel DMl enabled are taking several hours to even a day per table and are using high UNDO
I was checking and found that dbms_parallel_execute can be good solution. Can someone update if they had used for Big Update and can share his sample code to try
TIA
Subject: Big Update/DML
Sanjay
--
http://www.freelists.org/webpage/oracle-l
Received on Wed Aug 26 2020 - 16:30:31 CEST