Am 24.10.24 um 10:56 schrieb Stefan G. Weichinger via dovecot:
current assumption:
I mentioned adding a 2nd domain a few months ago.
This was done by editing stuff in postfixadmin and also changed things around the transports used.
Now I assume that I don't use lmtp anymore somehow ...
this is from mail.log. I send a testmail from my domain oops.co.at to an account in the problematic server:
Oct 24 10:45:21 mail2 postfix/qmgr[2703634]: 1437115F6A3: from=<office@oops.co.at>, size=1648, nrcpt=1 (queue active) Oct 24 10:45:21 mail2 postfix/smtpd[3014992]: disconnect from co.oops.co.at[45.84.138.128] ehlo=2 starttls=1 mail=1 rcpt=1 data=1 quit=1 commands=7
Oct 24 10:45:21 mail2 postfix/virtual[3020642]: 1437115F6A3: to=<stefan.weichinger@my-tld.at>, relay=virtual, delay=1.8, delays=1.8/0.01/0/0, dsn=2.0.0, status=sent (delivered to maildir) Oct 24 10:45:21 mail2 postfix/qmgr[2703634]: 1437115F6A3: removed
The relay is "virtual" ... I am digging now if the edited sql-queries somehow changed the transport for this domain.
Maybe I am completely wrong. dunno yet.
that seems to be the issue
I changed the transport map only for my test box, and the vacation script in sieve is triggered.
checks around:
# postmap -q problem.tld mysql:/etc/postfix/sql/mysql_transport_maps.cf virtual
Now I am scared to overrule the transport for the whole domains ;-)
In postfix we have:
virtual_transport = lmtp:unix:private/dovecot-lmtp
That worked for years.
I will test with single mailboxes and go step by step ...