Good evening and thanks for your answer Thomas,
here you can see the debug log of the last transmits to the assp server,
really good for debugging:
>Feb-24-16 22:34:13 [Worker_1] <doing full
<DAxODk0ODAgMDAwMDAgbiAKMDAwMDE4[CR][LF]
OTYwNyAwMDAwMCBuIAowMDAwMjU4NzkwIDAwMDAwIG4gCjAwMDAyNTg4MTIgMDAwMDAgbiAKMDAw[CR][LF]
MDI1ODk5MSAwMDAwMCBuIAowMDAwMjU5MTE4IDAwMDAwIG4gCjAwMDA0MDQyNzIgMDAwMDAgbiAK[CR][LF]
MDAwMDQwNDI5NSAwMDAwMCBuIAowMDAwNDA0NDc0IDAwMDAwIG4gCjAwMDA0MDQ2MDEgMDAwMDAg[CR][LF]
biAKMDAwMDUyMjI4NSAwMDAwMCBuIAp0cmFpbGVyCjw8IC9TaXplIDI3Ci9JbmZvIDEgMCBSCi9S[CR][LF]
b290IDIgMCBSCi9JRCBbIDwxOTI2MWU4OGU3OWZhZmRhOGUxM2NkMTVhMmRhMjA0YT4KPDE5MjYx[CR][LF]
ZTg4ZTc5ZmFmZGE4ZTEzY2QxNWEyZGEyMDRhPiBdCj4+CnN0YXJ0eHJlZgo1MzE5MTAKJSVFT0YK[CR][LF]
[CR][LF]
--=_mixed 00412150C1257F63_=--[CR][LF]
.>
>Feb-24-16 22:34:13 [Worker_1] <Maillog
>Feb-24-16 22:34:13 [Worker_1] <whitebody
>Feb-24-16 22:34:13 [Worker_1] <MessageSizeOK
>Feb-24-16 22:34:13 [Worker_1] <sendquedata
>Feb-24-16 22:34:13 [Worker_1] <SMTPTraffic - read OK
>Feb-24-16 22:34:13 [Worker_1] <SMTPTraffic - process read
>Feb-24-16 22:34:13 [Worker_1] <doing full <[CR][LF]
>
>Feb-24-16 22:34:13 [Worker_1] <Maillog
>Feb-24-16 22:34:13 [Worker_1] <whitebody
>Feb-24-16 22:34:13 [Worker_1] <MessageSizeOK
>Feb-24-16 22:34:13 [Worker_1] <sendquedata
>Feb-24-16 22:39:07 [Worker_1] <SMTPTraffic - read OK
>Feb-24-16 22:39:07 [Worker_1] <SMTPTraffic - process read
>Feb-24-16 22:39:07 [Worker_1] <doing line <421 4.4.2
smtp-proxy.localdomain Error: timeout exceeded[CR][LF]
I am not an expert in reviewing this, but I would assume that assp is
missing the [CR][LF].[CR][LF] which is splitted into two blocks. You see
any option to improve this, or am I wrong?
Best regards,
Johannes
>Can this splitting be a problem or bug within assp or postfix?
This should be normaly no problem - but it can be one. TCP is not
controlled by assp, this is done by the OS. The TCP packet size depends on
the MTU. It is not possible to control the OS-TCP-sendbuffer and the
TCP/IP package management from the application layer of the OSI model.
If postfix is not doing a readline or getline - instead using a sysread
and does not compose the content together to check for the
<CR><LF>.<CR><LF> - it will not detect it.
Currently assp writes as much as possible to the OS socket (in to the
TCP-sendbuffer). So it may possible that the syswrite operation of assp is
splited by the OS anywhere in the <CR><LF>.<CR><LF> . An typical
TCP-sendbuffer size is 16K or 64K. So even if assp would send the
<CR><LF>.<CR><LF> in a separate syswrite operation, there is no guaranty,
that this operation leads in to a separate TCP packet at the OS site.
To prevent such a missdetection, assp does a sysread but emulates a
getline until MaxBytes are received and switches back from bulk operation
to the getline mode at the end of the mail.
I don't know if postfix uses a similar logic.
Thomas
------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
_______________________________________________
Assp-user mailing list
Assp-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/assp-user
