I recently started getting "frozen" messages when trying to connect to a
particular mail server.
Previously I had had no trouble, but I believe the mail server for the
particular domain has just changed.
The relevant part of the debug message shows:
====================================================
Connecting to smtp.us.messagingengine.com [66.111.4.20]:25 ... connected
read response data: size=36
SMTP<< 220 mail.messagingengine.com ESMTP
SMTP>> EHLO myHostName
read response data: size=155
SMTP<< 250-mail.messagingengine.com
250-PIPELINING
250-SIZE 10120000
250-ETRN
250-STARTTLS
250-AUTH LOGIN PLAIN OTP
250-AUTH=LOGIN PLAIN OTP
250 8BITMIME
66.111.4.20 in hosts_avoid_tls? no (option unset)
SMTP>> STARTTLS
read response data: size=24
SMTP<< 220 Ready to start TLS
Initialized TLS
Calling SSL_connect
SSL_connect succeeded
reading pipe for subprocess 75562 (not ended)
read() yielded 0
remote delivery process 75562 ended: status=000b
set_process_info: 75561 delivering 19YExa-000Ptt-9Z
post-process me@??? (1)
LOG: MAIN
== me@??? R=dnslookup T=remote_smtp defer (-1):
smtp transport process returned non-zero status 0x000b: terminated by
signal 11
>>>>>>>>>>>>>>>> deliveries are done >>>>>>>>>>>>>>>>
=================================================
I added the line
hosts_avoid_tls=*
to my remote_smtp transport in the exim configure file and this cured
the problem.
I think what has happened here is that previously the receiving smtp
server didn't advertise TLS
capability, so my server connected without SSL. Now, on the new server,
it does advertise TLS
capability so my server tries to initiate an SSL link which fails
because I have not set up my OpenSSL correctly.
Has anyone else come across this behaviour? And what exactly is signal 11?
TIA
Graeme
