I have a stage that deletes 145 records from an 8 million row table. When I execute one delete from oracle developer it uses an index and takes less than a second. The datastage job does full table scans and takes over 4 minutes.
I am looking for a clue short of generating a sql profile to get the data stage job to use the index and run an efficient delete.
Oracle Connector Delete Statement Full Table Scan
Moderators: chulett, rschirm, roy
-
- Participant
- Posts: 3337
- Joined: Mon Jan 17, 2005 4:49 am
- Location: United Kingdom
i know this post is nearly a year old, but i just thought i'd mention for others looking for answers...
if the key col datatype is inconsistent with the index datatype in the database (eg double vs integer etc), oracle may well choose not to use the index - even though the predicate may be successful in locating the target rows.
so it may not affect your results but may affect the performance.
hope this helps,
steve
if the key col datatype is inconsistent with the index datatype in the database (eg double vs integer etc), oracle may well choose not to use the index - even though the predicate may be successful in locating the target rows.
so it may not affect your results but may affect the performance.
hope this helps,
steve