On Fri, 7 Apr 2006, Casey Allen Shobe wrote: > > b.www hosts websites for several clients. Among these is one who sends out a > mailing to a couple thousand people every day, using a commercial PHP-based > mailing list management program (*sigh*) called 1-2-all > (http://www.activecampaign.com/12all/). With the (fairly standard) qmail > setup, everything ran quietly and fine. After switching to exim, when the > mails are sent out each midday, load on b.www rises above 5, and load on > a.msa rises above 16. These numbers are consistant from day to day. > > What am I doing wrong?
You probably want to set your queue_only_load to be quite low, to stop exim trying immediate deliveries when the machine is busy. You might want to experiment with a two-stage queue runner (exim -qq) to see if that drains the queue faster than a standard queue runner. > That may keep the load from spiking so much on a.msa, but the mystery > remains - why does the load go up at all on b.www? That makes no sense to > me. Disk load. Exim is not particularly heavily optimised in this area. Try no_message_logs, split_spool_directory and also putting spool/exim/db on a ram disk and putting Exim's logs on a different disk from its spool. Tony. -- <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> http://dotat.at/ ${sg{\N${sg{\ N\}{([^N]*)(.)(.)(.*)}{\$1\$3\$2\$1\$3\n\$2\$3\$4\$3\n\$3\$2\$4}}\ \N}{([^N]*)(.)(.)(.*)}{\$1\$3\$2\$1\$3\n\$2\$3\$4\$3\n\$3\$2\$4}} -- ## List details at http://www.exim.org/mailman/listinfo/exim-users ## Exim details at http://www.exim.org/ ## Please use the Wiki with this list - http://www.exim.org/eximwiki/
