[QFJ-946] Improve misleading error message during Logon Created: 21/Mar/18 Updated: 26/Mar/18 |
|
Status: | Open |
Project: | QuickFIX/J |
Component/s: | Engine |
Affects Version/s: | 1.6.4 |
Fix Version/s: | None |
Type: | Improvement | Priority: | Default |
Reporter: | Nikolai Kulakov | Assignee: | Unassigned |
Resolution: | Unresolved | Votes: | 0 |
Labels: | None | ||
Environment: |
Linux |
Description |
It seems that the network connection between my company and our partner is very fast and the QuickFix/J engine suppose that their answering Logon message is coming before we send our Logon message: 8=FIX.4.2^A9=86^A35=A^A34=1^A49=fxw2^A52=20180321-14:09:01.447^A56=JLQD^A98=0^A108=30^A141=Y^A554=password^A10=253^A <2018-03-21 10:09:00.329 So, I do not see any way to establish the connection. |
Comments |
Comment by Nikolai Kulakov [ 21/Mar/18 ] |
After further investigation of the problem, I guess that the problem is not in the speed of the connection, but in the fact that we set ResetSeqNumFlag=Y in the our Logon message but in the answering Logon message it is not presented (so, the priority of the problem can be lowered). The code that is generating this error is in the method quickfix.Session#nextLogon: // Check for proper sequence reset response if (state.isResetSent() && !state.isResetReceived()) { disconnect("Received logon response before sending request", true); } It can be a problem indeed, but the disconnect message seems totally misleading for me. |
Comment by Christoph John [ 21/Mar/18 ] |
I don't know but what sense does it make to reset the sequence numbers only on one side? So I guess the check might have a point. Unless I'm missing something. |
Comment by Nikolai Kulakov [ 21/Mar/18 ] |
Just a real situation: our partner is ready to reset the sequence number on the one side only. It seems that using ResetSeqNumFlag for this purpose it is not nice by FIX standards, but also it is a fact that this situation is not related with the "Received logon response before sending request". |
Comment by Christoph John [ 21/Mar/18 ] |
I think it does not make sense to open a bug ticket for every counterparty that does not conform to normal FIX standard. IMHO the best would be to direct such questions to the mailing list. Please post follow up questions and your answers there. Thanks |
Comment by Nikolai Kulakov [ 21/Mar/18 ] |
John, I agree that it is no sence to open a bug ticket for a counetrparty that does not conform regular FIX standard (of course, I'll discuss this situation with them directly). But the ticket is about another thing and I'm really sorry if I'm describing it not clear. |