i've been struggling with this issue for almost a week now and after reading half of the forum, postfix- and amavis-docs i'm finally running out of ideas.
My setup is as follows:
- Linux box on the internet with official MX-records running the mta,antivir,antispam,logger
- Linux box behind firewall in corporate network running mailbox
E-Mail sending and receiving works fine (roughly 1000-3000 emails per day) through web ui as well as pop/imap/smtp from external.
Unhappily, after a varying amount of time the "deferred" queue starts filling up. The amount of emails being deferred varies - sometimes only 1-2, sometimes a few hundreds within a few hours.
here's the output from zimbra.log on such an occasion:
Nov 10 00:55:13 vs151138 postfix/smtp: C4B246600956: to=
, relay=127.0.0.1[127.0.0.1], delay=315, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending RCPT TO)
In the message queue view in the admin-ui the deferred emails are listed properly.
When I requeue the deferred emails or simply flush, they are being delivered properly but as mentioned above, after some time - varying from minutes to hours - they usually start being deferred again. In the admin-ui, the column "content-filter" always contains "smtp-amavis:[127.0.0.1]:10024".
Here's what i've tried so far:
- zmcontrol stop / start
- rebooted the box
- reinstalled zimbra on the box on the net
- increased the maxproc for amavis in postfix master.cf
I don't understand:
a) why they are being deferred at all
b) why they aren't reprocessed after 1000s since I have neither changed the postfix minimal_backoff_time nor the queue_run_delay - anyway they aren't even reprocessed after 8 hours.
Any hints are highly appreciated, this is a real pain for us right now.