We have a server which acts as a backup MX for many domains. Some of
these domains no longer have lower valued MXes, causing the message to be
frozen:
2001-09-06 14:56:27 15ezdo-0003Wa-00 <= sloanj@???
H=mail.uk.psi.com (mail) [154.8.2.142] P=smtp S=334
2001-09-06 14:56:28 15ezdo-0003Wa-00 lowest numbered MX record points to
local host: domain.deleted
2001-09-06 14:56:28 15ezdo-0003Wa-00 == postmaster@???
R=lookuphost defer (-1): lowest numbered MX record points to local host
2001-09-06 14:56:28 15ezdo-0003Wa-00 Frozen
Fair enough.
Now I could change this behaviour by setting self=fail (or self=pass) on
the appropriate router I guess.
However it seems that this frozen message is handled differently from
other frozen messages. I have set:
auto_thaw = 24h
ignore_errmsg_errors_after = 24h
Which seems to deal correctly with other frozen messages - they certainly
don't appear to persist in my queue past the end of my max retry timeout
of 7 days.
However the frozen messages of the type above are persisting. I have some
in my queue from more than 40 days ago.
Should these messages not be handled by the ignore_errmsg_errors_after
parameter, or at the very least by auto_thaw and retry settings?
