On 7 Sep 2011, at 09:39, Nigel Metheringham wrote:
>
> On 7 Sep 2011, at 02:09, Always Learning wrote:
>
>> Mailing list emails sent to me show a 30 minute delay at Tahini.
>
> Finding it hard to care here. Mailing lists are not supposed to be
> a real time medium.
>
They're not, but users are generally used to fairly short lags, and short lags do make conversations flow better. For announcement lists, a 40 minute delay means nothing, but for a discussion list, regular 40 minute delays would be unfortunate.
At Sussex, *all* our mailing lists are VERPd, and by Mailman not Exim. We did see problems where occasional use of a very large list would delay messages on all lists, by flooding the outgoing queue runner.
We mitigated delays by using a higher number of Mailman outgoing queue runners. We have this in Mailman/mm_cfg.py:
QRUNNERS = [
('ArchRunner', 4), # messages for the archiver
('BounceRunner', 4), # for processing the qfile/bounces directory
('CommandRunner', 4), # commands and bounces from the outside world
('IncomingRunner', 4), # posts from the outside world
('NewsRunner', 4), # outgoing messages to the nntpd
('OutgoingRunner', 32), # outgoing messages to the smtpd
('VirginRunner', 4), # internally crafted (virgin birth) messages
('RetryRunner', 4), # retry temporarily failed deliveries
]
The 32 outgoing queue runners mean that slow recipient hosts aren't such a problem.
You can also set VERP_DELIVERY_INTERVAL to something other than 1, to reduce the VERP overhead.
--
Ian Eiloart
Postmaster, University of Sussex
+44 (0) 1273 87-3148
