| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
Home -> Community -> Usenet -> c.d.o.server -> Re: Monitoring parallelism
Richard Foote <Richard.Foote_at_bigpond.com> wrote:
>Hi Rick,
>
>Check out V$PQ_SESSTAT, V$PQ_SLAVE, V$PQ_TQSTAT, and V$PQ_SYSSTAT for
>details regarding what/if/how parallel operations were executed.
Thanks a lot, that was exactly the info I needed.
But now I got some parallel processes spawned on another DB. One can easily recognize them in the process list.
>And as previously mentioned, you must specify the parallel clause
>(either at the table definition or via a hint) for Oracle to start
>creating more shovels and passing then around appropriately.
It seems that you are right here but then I just don't understand what the parameters in the init.ora are for. I just found out that changing the degree of parallelism of a table would be enough to parallelize some operations on this table, without changing any init.ora parameter.
The problem I have now is that running in parallel, things runs *slower*, since Oracle is more in favor of full table scans. I need to tweak some parameter to discourage Oracle to do excessive scans.
Bye
Rick
Received on Mon Aug 19 2002 - 18:21:18 CDT
![]() |
![]() |