In the archives there's a message from Jirí Smejkal
(
http://www.exim.org/mailman/htdig/exim-users/Week-of-Mon-20040823/msg00090.
html) that seems to describe exactly the problem I am experiencing. Has
anyone discovered the cause or had any further thoughts on the problem?
I'm using Exim 4.43 with the Exiscan patch and PostgreSQL 7.4.6, all
compiled from sources on FreeBSD 5.1. On connecting to Exim on TCP port 25
(or even on stdin using the -bh option) Exim emits the welcome banner and
then immediately closes the connection with the message '421 <hostname> lost
input connection'.
Example output from 'exim -d -bh 127.0.0.1':
host in host_reject_connection? no (option unset)
host in sender_unqualified_hosts? no (option unset)
host in recipient_unqualified_hosts? no (option unset)
host in helo_verify_hosts? no (option unset)
host in helo_try_verify_hosts? no (option unset)
host in helo_accept_junk_hosts? no (option unset)
SMTP>> 220 Spitfire Technology Group SMTP Service. Ready.
220 Spitfire Technology Group SMTP Service. Ready.
smtp_setup_msg entered
SMTP>> 421 test.spitfireuk.net lost input connection
421 test.spitfireuk.net lost input connection
LOG: smtp_connection MAIN
SMTP connection from localhost [127.0.0.1] lost (error: Resource
temporarily unavailable)
search_tidyup called
>>>>>>>>>>>>>>>> Exim pid=74519 terminating with rc=0 >>>>>>>>>>>>>>>>
The problem goes away if I compile Exim without PostgreSQL support. Further
experimentation has shown the problem occurs if I compile Exim without
PostgreSQL support, but do force Exim to be linked with libpq. Perhaps there
is a library namespace clash or other strange interaction.
Any clues would be greatly appreciated.
Regards,
Alastair Varnals.
Spitfire Technology Group.
