-----Original Message----- From: Mikael Abrahamsson
I am on mailing lists with many hundreds of people at other places and normal delivery time there is sub-minute. I see no technical reason today why an important list like nanog-l should reside on a mailing list platform where 10+ minute delivery time is reached in 20+ % of the cases and reached immediately when there is any load or moderate usage, and sub-5-minute delivery is reached in less than 20% of the cases.
my $.02... IMHO, I know of no out-of-the-box MTA that defaults to handle mailinglist loads. I run a mailinglist with ~750 members and ~25 posts per day. This server is a 300Mhz Cobalt RaQ3i w/ 256Mb of memory on a 256Kbs pipe (also running Apache w/ a fairly busy website: flagspot.net). Email turns around in the <5 minute time frame, and usually a heck of a lot quicker. Attached is a graph from yesterday showing the number of running Sendmail processes (green), and the number of emails in mailq (blue). Notice how there are very few emails that make it into the mailq in time to be polled by MRTG, most of these involve delivery issues with the remote mailserver (delays,timeouts,etc.). The trick to speedy turnaround is in the MTA configuration. This is chiefly achieved by using multiple queues, as well as severely reducing the timeouts such that email being sent to unreachable/overloaded mailhosts are quickly moved to the bottom of the delivery stack. -Jim P.
