The error that appears in my log is:
ERR No Data (Winsock error #11004)
The log is generated by the first routine ("if(Error)"), I presume when
a command has caused an error due to the network being interrupted.
I would presume that the second routine (the unknown RqType) doesn't
fire at all as that would be generated from a valid (but unknown)
command being completed correctly.
Once this occurs, the normal Connect command called in the next cycle
permanently fails until the application is restarted.
Dave
Wilfried Mestdagh wrote:
>Hello David,
>
>Do you have logs indicating which one cause the break, and what error /
>RqType it is ?
>
>I think in both cases you should call Abort();
>
>---
>Rgds, Wilfried [TeamICS]
>http://www.overbyte.be/eng/overbyte/teamics.html
>http://www.mestdagh.biz
>
>Monday, November 14, 2005, 10:33, David Lewis wrote:
>
>
>
>>I've got an application that I wrote a while ago now that is starting to
>>have some issues I need to look into.
>>
>>
>
>
>
>>Basically, this application contains a process that regularly polls for
>>emails (every 5 minutes) and gathers any available, stripping off up to
>>60 jpg attachements and processes them. (emails are from CCTV cameras)
>>
>>
>
>
>
>>The email server is on a network, and not connected via the 'net,
>>although the connection does go through a megastream link, which is
>>starting to become a little erratic, which I think is the cause of these
>>problems.
>>
>>
>
>
>
>>Normally, the application works fine, collects emails, logs everything
>>then sleeps until the next cycle and carries on 24/7
>>
>>
>
>
>
>>However, I think that occasionally, during the email collection routine,
>>the megastream link goes down briefly causing the routine to fail. At
>>this point, I log a fail, but then I cannot reconnect to the email
>>server again until I restart the application.
>>
>>
>
>
>
>>So, all I can imagine is that I am not doing the right thing when an
>>error occurs in the email retrieve routines.
>>
>>
>
>
>
>>In my "co_Pop3ClientRequestDone" routine, I have 2 error
>>handling/logging parts. The first being:
>>
>>
>
>
>
>>if(Error) // Error is passed to ClientRequestDone event
>> {
>> co_LogReport->Lines->Add(DateTimeToStr(Now( )) + " Email check
>>failed (" + IntToStr(RqType) + "/" + IntToStr(Error) + ")");
>> at_EmailError = true;
>> co_Pop3Client->Abort( );
>> return;
>> }
>>
>>
>
>
>
>>The second being:
>>
>>
>
>
>
>> switch(RqType)
>> {
>>...
>>...
>> default: // Handle unknown request and close
>> co_LogReport->Lines->Add(DateTimeToStr(Now( )) + " Email
>>check failed (" + IntToStr(RqType) + "/" + IntToStr(Error) + ")");
>> me_Exec(co_Pop3Client->Quit, "Quit");
>> break;
>> }
>>
>>
>
>
>
>>Now, the first uses Abort and this may not be the right way to close the
>>connection down, but I didn't want to create an infiniate loop where a
>>Quit command called this routine with and error and then everything
>>falls over.
>>
>>
>
>
>
>>What is the 'right' way to close the connection down in such a way that
>>the next attempt will be a 'clean' one? I presume I haven't closed
>>everything down properly which is causing be to fail on the next connection.
>>
>>
>
>
>
>>At the moment, unfortunately, this software is 'on-site' and not in a
>>good position to test/debug so I would like to be confident that any
>>changes will be in the right direction.
>>
>>
>
>
>
>
>>Thanks,
>>
>>
>
>
>
>>Dave
>>
>>
>
>
>
--
To unsubscribe or change your settings for TWSocket mailing list
please goto http://www.elists.org/mailman/listinfo/twsocket
Visit our website at http://www.overbyte.be
