We run a 90000+ subscribers newsletter type mailing list on a dedicated
server. I have taken this opportunity to gather information on delivery
behaviour/speed using different concurrencyremote settings (150, 250 and 500)
and graph the results.
The result of the comparison is rather astonishing (for me ;-) as there
is not really a big difference.
The "main" work at concurrencyremote=500 was finished after about 1250
seconds, at concurrencyremote=150 it was finished after about 1450
seconds; concurrencyremote=250 is in between at about 1350 seconds.
The number of finished successful deliveries/second is nearly the same
for all three data sets (about 75-80 deliveries/second).
However the number of failures/deferrals per second was lower in the
150 data set than in the 250 and much lower than in the 500.
Also the maximum and median delivery times were smaller for the 150 set.
(as the list is ezmlm maintained by far the most "failures" are deferrals).
*MY* conclusion from that comparisons is that the power of the
qmail-bigconcurrency patch is probably commonly overestimated
and the patch is kinda useless.
PLEASE NOTE: the data sets are collected from delivery cycles of three
successive weeks (the newsletter is a weekly one). Although it's
delivered the same weekday (Friday) and around the same time (early
afternoon GMT+2) the load on the remote (i.e. receiving) mail servers
has a large impact on the data. This is even more true as 90% of
the messages are sent to only 300 unique IP addresses (some of which
are surely hidden behind load balancers).
Thus minor tendencies are to be handled with care and the data sets
may not be really representative.
I have set up a WebPage at
http://www.lamer.de/maex/creative/software/qmail/deliveries/
which contains a bit more explanations and a lot (about 20) of lessened
images (full size by clicking on the small images). Although the images
in the page are lessened the whole page has about 300 KB, so it may take
a while to load completely.
All the graphs were made with the help of the qmail logfile, perl, awk,
grep and gnuplot ;-)
Thanks to Peter van Dijk for his comments and thoughts while previewing
the weekly results.
I'd be very interested in your opinions/comments.
\Maex
--
SpaceNet AG | Joseph-Dollinger-Bogen 14 | Fon: +49 (89) 32356-0
Research & Development | D-80807 Muenchen | Fax: +49 (89) 32356-299
Stress is when you wake up screaming and you realize you haven't fallen
asleep yet.
