Hi all,
we've just started observing strange malfunctioning of Exim right after
upgrade from v4.95 to v4.96 on our centos7 and rockylinux8 servers.
Incoming emails are not delivered at once, but instead Exim logs one the
following error messages into mainlog and defers the email in exim
queue. After some time (minutes, sometimes hours) it delivers the email.
On MX servers, we observe the following error extremely frequently for
many domains:
== rcpt@??? R=xxx T=remote_smtp_envelope_to_add defer (-54): retry
time not reached for any host for 'xyz.com'
Sometimes we see also this error:
== rcpt@??? R=xxx (-36): host lookup for <imap_server> did not
complete (DNS timeout?)
There is no problem with DNS resolution of the relevant domains on those
MX servers whatsover, judging by using dig and host command in shell.
On IMAP servers receiving emails from the above MX servers, the new and
very frequent error is now this (even when the mid corresponds to just
one recipient):
<mid> no immediate delivery: more than 10 messages received in one
connection
Exim leaves such email in its queue for few minutes and then it
eventually delivers it.
The only way to speed up exim to its previous standard behaviour is to
push the message delivery forward manually via the exim -M <mid>
command, on both MX and IMAP servers.
Is it possible exim 4.96 changed the previous default behaviour and is
now sending many more emails in one connection than before? We tried
settings
smtp_accept_max_per_connection = 100
smtp_accept_queue_per_connection = 100
in exim.conf, but to no avail.
Any ideas what may be causing this strange behaviour after upgrade to 4.96?
--
S pozdravem
Ján Lalinský
Webhosting C4
ČESKÝ WEBHOSTING s.r.o.
Tel: +420 222 292 444
E-mail: info@???
http://www.c4.cz
