Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: V$SESSION_WAIT : event question
Billy,
You hit the nail right on the head. One way to check your theory is to
inspect the serial number reported in v$session_wait. That number
changes each time the session goes into a new wait state. If it is
increasing, that means problems on the client side.
One common reason you will see the session wait on this event is when
the client is requesting many rows ONE AT A TIME, rather than by using
an array fetch. This would show up with the same SQL being executed by
the session. In this scenario, Oracle will return the row SO quickly
that the session will always appear to be waiting ofr the client. Also,
check the WAIT_TIME column. It needs to be 0 (meaning 'currently
waiting')
Ask your developer if they are using array fetch.......
James
In article <8104kg$63k$1_at_ctb-nnrp1.saix.net>,
"Billy Verreynne" <vslabs_at_onwe.co.za> wrote:
> Anyone has an idea what this events is about? :
>
> Event : SQL*Net message from client
> Param1: Driver ID
> Param2: number of bytes or blocks
>
> I have a process running, still being shown as active, which the
users says
> is now taking forever, and even though rows are being processed
(according
> to session stats), the majority of the events waited on, is this one.
>
> I assume that "message from client" is just that - Oracle gets a row,
> returns it to the client which in turns processes it, before
requesting the
> next one. This means that from an Oracle Server perspective, the
majority of
> its time for that session is spend waiting for the client to request
more
> data. So the performance problem, if there is one, has then to do
with how
> fast/well the client actually processes the data. Is this assumption
valid?
>
> Comments will be much appreciated.
>
> regards,
> Billy
>
>
Sent via Deja.com http://www.deja.com/
Before you buy.
Received on Thu Nov 18 1999 - 03:26:48 CST