Hi guys
Our spool is full of undeliverable bounce mail.
We have the parameters
ignore_bounce_errors_after = 2h
timeout_frozen_after=1d
but we have a lot of bounce mail up to 3 days old and only 3 days
because we did a big cleanout on Friday. Here's an example for one 46h
old
# /opt/exim/bin/exim -v -M 1F0Gq9-0001zw-HC
delivering 1F0Gq9-0001zw-HC
Connecting to relay.verizon.net [206.46.232.11]:25 ... connected
SMTP<< 220 sv3pub.verizon.net MailPass SMTP server v1.2.0 -
112105154401JY+PrW ready Mon, 23 Jan 2006 04:15:11 -0600
SMTP>> EHLO mx2.bt.net
SMTP<< 250-sv3pub.verizon.net
250-8BITMIME
250 SIZE 20971520
SMTP>> MAIL FROM:<> SIZE=2404
SMTP<< 250 Sender <> OK
SMTP>> RCPT TO:<brandonyc@???>
SMTP<< 550 5.1.1 unknown or illegal alias: brandonyc@???
SMTP>> QUIT
LOG: MAIN
** brandonyc@??? R=lookuphost T=smtp: SMTP error from remote
mailer after RCPT TO:<brandonyc@???>: host relay.verizon.net
[206.46.232.11]: 550 5.1.1 unknown or illegal alias:
brandonyc@???
LOG: MAIN
brandonyc@???: error ignored
LOG: MAIN
Completed
# cat /var/spool/exim/input/9/1F0Gq9-0001zw-HC-H
cat: /var/spool/exim/input/9/1F0Gq9-0001zw-HC-H: No such file or
directory
In my opinion, the first time this occurred, the bounce should have been
frozen as this address in undeliverable however as the above shows, the
message hadn't been frozen. Then the 'ignore_bounce_errors_after'
parameter should have caused its deletion after 2h. Why did this not
happen? We had 120k similar bounces messages in the queue this morning
and I'd like to get rid of these automatically? After the firced
delivery the message was deleted as it had been on the queue for longer
than ignore_bounces_errors_after.
Am I doing something wrong here?
Thanks
Martin
