Re: Small addition

vafanassiev_at_aapt.com.au (Vsevolod Afanassiev) wrote in message news:<4f7d504c.0302042321.2251719b_at_posting.google.com>...
> Additions:
> 1. I increased JAVA_POOL_SIZE to 100 MB, SHARED_POOL_SIZE to 160 MB,
> DB_BUFFERS to 30,000 - this had no effect.
> 2. I monitored session that was doing 'alter java class ... resolve'
> via V$SESSION_EVENT - it had small number of following waits
>
> undo segment extension
> log buffer space
> log file switch completion
> log file sync
>
> I created two large (500 MB) rollback segments without OPTIMAL
> and took existing rollback segments offline. I also increased
> the number of online redo groups to 9 and size of online redo logs
> to 50M, and LOG_BUFFER to 1.6 MB. This led to a small improvement...
>
> Main wait events are
>
> SQL*Net message to client
> SQL*Net message from client

Wait event SQL*Net message from client is OK. That's main wait event on all systems, that's the time spend by users looking on monitor and trying to understand what he just saw.
However, if SQL*Net message TO client is one of the main events you should definitely take a close look on your network environment, because this event should be very small on almost all systems.

with regards

--
_________________________________________

Dusan Bolek, Ing.
Oracle team leader

Note: pagesflames_at_usa.net has been cancelled due to changes (maybe we
can call it an overture to bankruptcy) on that server. I'm still using
this email to prevent SPAM. Maybe one day I will change it and have a
proper mail even for news, but right now I can be reached by this
email.