Too Many oracle ora_p0?? processes in Solaris 10

Question

I have a very intense update process which has about 8 SQL query - every one of them is having the HINT "NO_PARALLEL" as well as the stmt "Alter session disable dml;" before we execute the query.
Even the Index is also hinted as no parallel.
Every time this process run, i see number of ora_p??? process comes up and it never went goes down.
No when ever this happens, I am seeing lot of slowness in my transaction times. Please help and advice....

Answer 1

1. Oracle 10g 10.2.0
SQL*Plus: Release 10.2.0.2.0
2. Yes, solaris 10g has been configured for this.
3. We have 4 quad procs
4. Plan
<PRE>
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | Pstart| Pstop| TQ |IN-OUT| PQ Distrib |
| 0 | DELETE STATEMENT | | 131K| 11M| 52285 (1)| 00:15:42 | | | | | |
| 1 | DELETE | TABLE1 | | | | | | | | | |
| 2 | PX COORDINATOR | | | | | | | | | | |
| 3 | PX SEND QC (RANDOM) | :TQ10000 | 131K| 11M| 52285 (1)| 00:15:42 | | | Q1,00 | P->S | QC (RAND) |
| 4 | PX PARTITION RANGE ALL | | 131K| 11M| 52285 (1)| 00:15:42 | 1 | 9| Q1,00 | PCWC | |
| 5 | INLIST ITERATOR | | | | | | | | Q1,00 | PCWC | |
|* 6 | TABLE ACCESS BY LOCAL INDEX ROWID| TABLE1 | 131K| 11M| 52285 (1)| 00:15:42 | 1 | 9| Q1,00 | PCWP | |
|* 7 | INDEX RANGE SCAN | IDX_ID_1 | 218K| | 196 (2)| 00:00:04 | 1 | 9| Q1,00 | PCWP | |
</PRE>
Predicate Information (identified by operation id):
6 - filter("COL1"='ABC')
7 - access("ID_1"='01234567891092' OR "ID_1" = "0000001233445566')
22 rows selected.
Message was edited by:
user608665