[Mailman-Users] mailman not functional

[Mark Sapiro]

On 10/17/19 3:59 PM, Steven Jones wrote:
> thanks
> 
> I am sort of suspecting postfix as /var/log/maillog is not being written to.
> 
> meanwhile,
> 
> /var/log/mailman/smtp has something odd. The time to send email varies between 0.017 and 180 seconds plus,  and its getting worse,
> 
> ==========
> Oct 18 11:26:13 2019 (9398) <mailman.0.1571274477.9729.postgradcoordinators at lists.vuw.ac.nz> smtp to postgradcoordinators for 1 recips, completed in 0.017 seconds
> Oct 18 11:26:26 2019 (9398) <mailman.1.1571274477.9729.postgradcoordinators at lists.vuw.ac.nz> smtp to postgradcoordinators for 14 recips, completed in 12.419 seconds
> Oct 18 11:26:27 2019 (9398) <mailman.0.1571274594.9753.vuw-schooladmin at lists.vuw.ac.nz> smtp to vuw-schooladmin for 14 recips, completed in 0.029 seconds
> Oct 18 11:26:27 2019 (9398) n/a smtp to its-alerts for 10 recips, completed in 0.073 seconds
> Oct 18 11:26:27 2019 (9398) <mailman.0.1571274783.10156.postgradcoordinators at lists.vuw.ac.nz> smtp to postgradcoordinators for 1 recips, completed in 0.019 seconds
> Oct 18 11:26:27 2019 (9398) <mailman.1.1571274783.10156.postgradcoordinators at lists.vuw.ac.nz> smtp to postgradcoordinators for 14 recips, completed in 0.019 seconds
> Oct 18 11:26:39 2019 (9398) n/a smtp to its-alerts for 10 recips, completed in 10.111 seconds
> Oct 18 11:26:41 2019 (9398)  <ME2PR01MB4292894A699E1DF46C8DABE8C46D0 at ME2PR01MB4292.ausprd01.prod.outlook.com> smtp to fad-design-postgrads for 102 recips, completed in 0.248 seconds
> Oct 18 11:26:41 2019 (9398) n/a smtp to its-alerts for 10 recips, completed in 0.024 seconds
> Oct 18 11:26:41 2019 (9398)  <ME2PR01MB4292894A699E1DF46C8DABE8C46D0 at ME2PR01MB4292.ausprd01.prod.outlook.com> smtp to fad-design for 56 recips, completed in 0.030 seconds
> Oct 18 11:29:45 2019 (9398) <SYAPR01MB26875792826537251B8585DDD46D0 at SYAPR01MB2687.ausprd01.prod.outlook.com> smtp to nz-libs for 1844 recips, completed in 184.400 seconds
> Oct 18 11:31:31 2019 (9398) <mailman.10788.1571276386.12225.teachingandlearning at lists.vuw.ac.nz> smtp to teachingandlearning for 1 recips, completed in 103.629 seconds
> Oct 18 11:34:22 2019 (9398) n/a smtp to its-alerts for 10 recips, completed in 169.826 seconds
> Oct 18 11:36:26 2019 (9398) n/a smtp to its-alerts for 10 recips, completed in 123.446 seconds
> Oct 18 11:39:33 2019 (9398)  <ME2PR01MB429218FD19F84FE6C16AA291C46D0 at ME2PR01MB4292.ausprd01.prod.outlook.com> smtp to fad-design for 56 recips, completed in 187.700 seconds
> 8><----


Mailman's out queue is backlogged. See
<https://wiki.list.org/x/17892002>. Something is changed affecting
Postfix. Possibly you had a local DNS cache that is now not working and
Postfix DNS lookups are taking a long time. Also consider a separate
Postfix port for Mailman delivery with minimal checking. On
mail.python.org we use something like this in master.cf

127.0.0.1:8027
          inet  n       -       -       -        -      smtpd
        -o smtpd_recipient_restrictions=permit_mynetworks,reject
        -o smtpd_client_restrictions=
        -o smtpd_helo_restrictions=
        -o smtpd_sender_restrictions=
        -o smtpd_data_restrictions=

...
> =========
> [root at vuwunicomailmn1 mailman]# tail qrunner
> Oct 18 09:32:13 2019 (9400) RetryRunner qrunner started.
> Oct 18 09:32:13 2019 (9395) CommandRunner qrunner started.
> Oct 18 09:32:13 2019 (9396) IncomingRunner qrunner started.
> Oct 18 09:32:13 2019 (9399) VirginRunner qrunner started.
> Oct 18 09:32:13 2019 (9397) NewsRunner qrunner started.
> Oct 18 09:32:13 2019 (9394) BounceRunner qrunner started.
> Oct 18 09:32:13 2019 (9398) OutgoingRunner qrunner started.
> Oct 18 09:32:13 2019 (8205) OutgoingRunner qrunner exiting.
> Oct 18 09:32:13 2019 (8172) Master qrunner detected subprocess exit
> (pid: 8205, sig: None, sts: 15, class: OutgoingRunner, slice: 1/1)
> [root at vuwunicomailmn1 mailman]#
> =========
> 
> and its now 11:50am
> 
> Ok to assume the above in the qrunner log is normal?
>


The last three lines are not normal OutgoingRunner has died, but it
seems these are the result of a Mailman restart and the old
OutgoingRunner persisted for a while because it had to finish delivery
of the current message. Note the out of sequence PID and the fact that
ps shows 9398 is running.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan