Updates to a table with post-update-triggers out of a pl/sql loop get slower through the loop
We encountered a strange phenomenan, but didn't find a corresponding bug. Currently we cannot open any SR's due to administrative troubles, so I describe the problem here. Perhaps somebody encountered the same problem (we cannot beleive this didn't happen quite frequently)
Problem:
Multiple updates to a given table with post-update-triggers out of a pl/sql loop lead to increasing uga consumption by that session, and increasing cpu-times for these updates.
After the pl/sql procedure ends, uga remains reserved. The memory is freed only by the next top-level statement issued by the session.
We are on 11.1.0.6 (yes, desupported, but 11.1.0.7 is not an option due to a bug with triggers and returning dml).
Problem:
Multiple updates to a given table with post-update-triggers out of a pl/sql loop lead to increasing uga consumption by that session, and increasing cpu-times for these updates.
After the pl/sql procedure ends, uga remains reserved. The memory is freed only by the next top-level statement issued by the session.
We are on 11.1.0.6 (yes, desupported, but 11.1.0.7 is not an option due to a bug with triggers and returning dml).
0