Re: ORA-00018 but where are the connections coming from?
Date: Tue, 22 Oct 2013 14:39:38 +0100
Message-ID: <CAPMSPxPjiJO8q+nkaF_PEh=WV4V3x9+ArVo8rZArxNq6xY1mpg_at_mail.gmail.com>
Oops
Spoke a bit too soon, I realise now that some of the jdbc programmes when they end are being recorded as "logoff by cleanup" and we do not have the patch that gets that information recorded in sys.aud$, the logoff audit only goes into a OS file in adump.
Still I think I am on the right lines.
Regards
Pete
On 22 October 2013 13:44, Peter Hitchman <pjhoraclel_at_gmail.com> wrote:
> Hi
> Thanks both for the replies.
> I got the kick I needed. The problem does appear to come down to
> applications for one app server all connecting from around mid-night and
> running for longer than expected, leading to the all the session slots
> being exhausted by the afternoon.
>
> Regards
> Pete
>
>
> On 21 October 2013 17:24, Andy Klock <andy_at_oracledepot.com> wrote:
>
>> On Mon, Oct 21, 2013 at 11:58 AM, Rich Jesse <
>> rjoralist3_at_society.servebeer.com> wrote:
>> >
>> > There's the possibility that the connections aren't logged via the
>> listener
>> > (bequeath connections do not use the listener) or you may have a large
>> > number of job processes available that all became active, but the
>> listener
>> > log could be a start for you.
>> >
>> > HTH! GL!
>> >
>> > Rich
>> >
>> >
>> >
>> Yeah, definitely the listener is a good place to start. But, also in 11g
>> auditing defaults to DB and audits CREATE SESSION, so if it hasn't been
>> disabled there may be some answers in DBA_AUDIT_SESSION:
>>
>> http://docs.oracle.com/cd/B28359_01/server.111/b28320/statviews_3071.htm
>>
>> Sometimes it's not new sessions that are the cause of the problem, but
>> rather sessions that are supposed to close are failing to do so.
>>
>> Andy
>>
>>
>> --
>> http://www.freelists.org/webpage/oracle-l
>>
>>
>>
>
>
> --
> Regards
>
> Pete
>
-- Regards Pete -- http://www.freelists.org/webpage/oracle-lReceived on Tue Oct 22 2013 - 15:39:38 CEST