-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Wed, 2 Jan 2013, DormitionSkete@hotmail.com wrote:
On Jan 2, 2013, at 11:12 AM, Andrzej A. Filip wrote:
On 01/02/2013 06:53 PM, DormitionSkete@hotmail.com wrote:
[...] This is the mail header from the mqueue:
[...] MDeferred: dovecot mailer (/usr/local/libexec/dovecot/dovecot-lda) exited with EX_TEMPFAIL [...]
Check dovecot's logs If it does not help the push delivery of messages ni sendmail's queue in verbose mode: "sendmail -Am -v -q"
Well, the logs showed this:
Note, sendmail started the /usr/local/libexec/dovecot/dovecot-lda, but the LDA failed temporarily. Where does Dovecot LDA log to?? I suppose that the LDA is executed with permissions, that does not allow to write to the mail storage or that the LDA does not have permission to access the userdb. Maybe it's the same problem as in the thread "From Sendmail to Dovecot mdbox, what is missing.".
I suggest to replace the "P=" in the Dovecot mailer with a wrapper script, you get all the necessary information with:
================ wraplda.sh #!/bin/sh
( date ; id echo "$@" /usr/local/libexec/dovecot/dovecot-lda "$@" rc=$? echo exit code: $rc exit $rc ) >> /tmp/lda.log 2>&1
Put it into your "smrsh" directory.
So I put this in my /etc/hosts file:
10.211.55.6 localhost
And I added DOVECOT to my /etc/hosts 127.0.0.1 line. This cleared the queue, but now all of the mail for the virtual users bounces with this:
No! That's against the idea of the placeholder "DOVECOT". You see that the sendmail part works, because the Dovecot LDA was started.
The original message was received at Wed, 2 Jan 2013 11:16:27 -0700 (MST) from openindiana [IPv6:::1]
----- The following addresses had permanent fatal errors ----- <hacwebstore@localhost> (reason: 550 5.1.1 User unknown) hacwebstore@DOVECOT (reason: 550 5.1.1 User unknown) (expanded from: <hacwebstore@localhost>)
----- Transcript of session follows ----- 550 5.1.1 hacwebstore@DOVECOT... User unknown 550 5.1.1 <hacwebstore@localhost>... User unknown
Because the LDA returns EX_TEMPFAIL, the messages stays in the queue until the temporary failure is fixed or the maximum queue stay time (usually 5 days) is reached. But in this case, the error message should be different, IMHO. I don't think that this error is from the latest test.
Regards,
Steffen Kaiser -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux)
iQEVAwUBUOUzYWoxLS8a3A9mAQL+WQgAtguL23X3Kqozj4rwakuRnx4kfEVwsnJK lvZXtJaj3AKEcQJVGhL41GfUJkvfvcucRmctU+srXDf2R47ZmcxTjpWnRZk8zsd2 x4R31lkHdZE0xXUnPIz/pjyjXv4A4I2Pv0Euyu3oI3m1yR7kENz5f/spgcC0A/3Z 7j0byshbTjac3212sAux5hI13B1bkec11tREaw2w6PVjYF6wsQ3PP9fFj0M7Ctmc RWazH6Sxe8dBSjW9Wd+SlYbhWl+j7A8fErxnOa34zSJQzXv1Dx5+nOrEyuDm7q7e A8ZWxc8Ss0ilXXSY0XUwvDNubQWm9wGUd/ajLt9h9Kg1Ade7zO5Dzg== =64Wz -----END PGP SIGNATURE-----