Re: [Quickfix-developers] Logon-Mystery Disconnect

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

I'm not ignoring your requests for better logging.  It's been noted and =20=

this will be improved.  Do keep in mind that the logging had been =20
greatly improved in 1.7.0, including events for many disconnection =20
scenarios.  This may shed light on your problem.  If you don't want to =20=

upgrade, you can at least do a diff and add the new logging related =20
stuff from the new session class.

Also, if you can post your configuration file, that would probably be =20=

helpful in trying to duplicate what you are seeing.

--oren

On Apr 29, 2004, at 8:49 AM, Howard S. Engelhart wrote:

> Thanks for the suggestion, but I don't think it's that.  I destroy my  =
=20
> FIX Application, Session, Log, etc objects and remove the logs on a =20=

> nightly basis, so unless there's an initialization problem when these =20=

> objects are created, I am pretty certain my sequence numbers are =20
> coming up 1,1 in the morning.  In the example below he comes in at =20
> SeqNum=3D1 and is disconnected.  4 minutes later he comes in at 2 and =
is =20
> allowed to log on, however qf logs that it expected SeqNum=3D1, =
received =20
> 2.  So it does not look to be a too low seq num issue.  Not to beat a =20=

> dead horse, but again, adding a reason indicator to the =20
> "Disconnecting" message would go a long way to clearing up, if nothing =
=20
> else, at least why qf thinks it should be disconnecting the session.
>
> -----Original Message-----
> From: Miller, Oren [mailto:OM...@ri...]
> Sent: Thursday, April 29, 2004 9:22 AM
> To: Howard S. Engelhart; qui...@li...
> Subject: Re: [Quickfix-developers] Logon-Mystery Disconnect
>
>
> This sounds like he may be logging in with a sequence number that is =20=

> too low.  What will happen is he will attempt logons, get disconnected =
=20
> due to the sequence number being to low.  Every time he is sending a =20=

> logon, he increments his sequence number until he eventually reaches =20=

> the right one.  Perhaps yor sequence numbers arn't being reset when =20=

> you think they are?
>
> --------------------------
> Sent from my BlackBerry Wireless Handheld
>
>
> -----Original Message-----
> From: Howard S. Engelhart <ho...@ex...>
> To: QuickFIX Questions (E-mail) =20
> <qui...@li...>
> Sent: Thu Apr 29 07:22:23 2004
> Subject: RE: [Quickfix-developers] Logon-Mystery Disconnect
>
> Yes.  The enigmatic "Disconnecting" especially.  This morning I =20
> observed a repeat of the exact same behaviour as yesterday.  The =20
> client attempts the logon at 12:05 gmt is disconnected for reasons =20
> that I cannot fathom from the logs, and then successfully reconnects =20=

> and logs on 4 minutes later.  If anyone has experience/insight into =20=

> this, it would be greatly appreciated.
>
> =
8=3DFIX.4.0`9=3D74`35=3DA`49=3DCLIENT`56=3DSERVER`34=3D1`50=3DADMIN`43=3DN=
`52=3D20040429-=20
> 12:05:16`98=3D0`108=3D60`10=3D238`
> =
8=3DFIX.4.0`9=3D74`35=3DA`49=3DCLIENT`56=3DSERVER`34=3D2`50=3DADMIN`43=3DN=
`52=3D20040429-=20
> 12:09:10`98=3D0`108=3D60`10=3D237`
> =
8=3DFIX.4.0`9=3D68`35=3D1`49=3DCLIENT`56=3DSERVER`34=3D3`50=3DADMIN`43=3DN=
`52=3D20040429-=20
> 12:09:10`112=3D3`10=3D202`
> =
8=3DFIX.4.0`9=3D95`35=3D4`49=3DCLIENT`56=3DSERVER`34=3D1`50=3DADMIN`43=3DY=
`52=3D20040429-=20
> 12:09:10`122=3D20040429-12:09:10`123=3DY`36=3D4`10=3D015`
> =
8=3DFIX.4.0`9=3D95`35=3D4`49=3DCLIENT`56=3DSERVER`34=3D1`50=3DADMIN`43=3DY=
`52=3D20040429-=20
> 12:09:10`122=3D20040429-12:09:10`123=3DY`36=3D4`10=3D015`
> =
8=3DFIX.4.0`9=3D68`35=3D1`49=3DCLIENT`56=3DSERVER`34=3D4`50=3DADMIN`43=3DN=
`52=3D20040429-=20
> 12:10:14`112=3D4`10=3D200`
>
> 20040429-11:30:03 : Created session
> 20040429-12:05:16 : Disconnecting
> 20040429-12:09:10 : Received logon request
> 20040429-12:09:10 : Responding to logon request
> 20040429-12:09:10 : MsgSeqNum too high RECEIVED: 2 EXPECTED: 1
> 20040429-12:09:10 : Sent ResendRequest FROM: 1 TO: 999999
> 20040429-12:09:10 : MsgSeqNum too high RECEIVED: 3 EXPECTED: 1
> 20040429-12:09:10 : Sent ResendRequest FROM: 1 TO: 999999
> 20040429-12:09:10 : Received SequenceReset FROM: 1 TO: 4
> 20040429-12:09:10 : MsgSeqNum too low RECEIVED: 1 EXPECTED: 4 PosDup: =
Y
>
> -----Original Message-----
> From: Joerg Thoennes [mailto:Joe...@ma...]
> Sent: Wednesday, April 28, 2004 10:00 AM
> To: Howard S. Engelhart
> Cc: QuickFIX Questions (E-mail)
> Subject: Re: [Quickfix-developers] Logon-Mystery Disconnect
>
>
>>> On a side note, it would help to have a bit more verbosity in the
>>> event logging, particuarly where disconnects are concerned.  I
>>> would gladly volunteer my time towards putting this in if I can
>>> help.
>
> Yes, QF could be a lot more verbose in the event log, esp. for the
> following cases:
>
> 1. Rejected logon
> 2. Disconnect
> 3. Logout with user provided logout reason
> 4. Dropped messages fragments which fail body length or CRC check.
>
> Cheers, J=F6rg
>
> --=20
> Joerg Thoennes
>                                 http://macd.com
> Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
> Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle =20=

> 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id149&alloc_id=9966&op=3Dick
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle =20=

> 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id149&alloc_id=9966&op=3Dclick
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>