okay, I'll bite -- What else is going on in this block? Are queries, potentially involving blocks and indexes affected by the inserts being iteratively queried between the inserts? Since the bind variables are constant in name, this implies they must be set to new values either by query or new value assignment in the pl/sql code. The former could be affected by what our friend Tom Kyte refers to as write consistency, while the latter would still imply a full stack line turn around to the client (that is where your pl/sql is running, right?). Sounds to me like either possibility is the wrong way to run a railroad, even if some bug can be discovered and fixed to make the performance within your service level requirements. Regards, mwf -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Baumgartel, Paul Sent: Wednesday, March 21, 2007 9:44 AM To: 'Alberto Dell'Era'; Baumgartel, Paul Cc: oracle-l@xxxxxxxxxxxxx Subject: RE: Increasing CPU on every execution of an anonymous block sent from WebLogic The timings are for serial execution of the block by the same connection. The text of the block, including bind names, is identical every time. Paul Baumgartel CREDIT SUISSE Information Technology DBA & Admin - NY, KIGA 1 One Madison Avenue New York, NY 10010 USA Phone 212.538.1143 paul.baumgartel@xxxxxxxxxxxxxxxxx www.credit-suisse.com -----Original Message----- From: Alberto Dell'Era [mailto:alberto.dellera@xxxxxxxxx] Sent: Tuesday, March 20, 2007 3:52 PM To: paul.baumgartel@xxxxxxxxxxxxxxxxx Cc: oracle-l@xxxxxxxxxxxxx Subject: Re: Increasing CPU on every execution of an anonymous block sent from WebLogic > Analysis of the trace files for the bind-variable version reveals that the > CPU time to EXEC the anonymous block (and, to a lesser extent, its > constituent INSERT statements) steadily increases. The first time > through the CPU time is about 0.02 second. By the last execution > before the timeout, the CPU time is just about 1.0 second. > Interestingly, the logical I/O stats for each execution are quite low > (<100 current reads and consistent reads) and quite consistent. Are those timings for a serial execution of the block by the same connection, or by concurrent execution by different connections ? Is the block identical for bind variables names also ? ie begin insert ... values (:bind_one); insert ... values (:bind_one); end; or begin insert ... values (:bind_one); insert ... values (:bind_TWO); end; -- Alberto Dell'Era "dulce bellum inexpertis" ============================================================================ == Please access the attached hyperlink for an important electronic communications disclaimer: http://www.credit-suisse.com/legal/en/disclaimer_email_ib.html ============================================================================ == -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l