I forward some mail from the machine "iron" to "corn." Recently, the
connection between the 2 got blocked. After awhile, every message to
corn immediately generated a DSN message ("Message ... has been
frozen"). I got a loop, as each DSN message produced a new one. There
were over 79,000 messages in the queue when I caught up with it.
I've recovered from that particular incident, but wonder if people have
suggestions for avoiding it in the future. The sequence was
1. exim on iron receives a message to be delivered to corn
2. exim on iron notes that corn is unreachable, and generates a DSN to
myself (via postmaster) on iron
3. exim on iron delivers that message to cyrus on iron by lmtp
4. cyrus consults its sieve file for me, and discovers I want to send a
copy of the message to corn
5. cyrus delivers the message to exim on iron, which causes the cycle to
repeat.
I realize that if the cyrus seive rule did not forward bounce messages
it would break the loop, but I wanted to receive them in general.
I'm not certain if it would make a difference if I used a .forward file
on iron to do the mail handling (which would keep it in exim), rather
than having cyrus handle the mail with seive. I know exim tries to
avoid mail loops; I'm not sure how possible that it when the loops are
generating new messages.
Any thoughts?
Ross
--
## List details at http://lists.exim.org/mailman/listinfo/exim-users
## Exim details at http://www.exim.org/
## Please use the Wiki with this list - http://wiki.exim.org/
