Not staggeringly large (or small) ones, no...
It often happens that Exim's periodic queue runs become synchronised
on different hosts. These may be multiple servers for a particular mail
domain, or just ones that share common infrastructure (e.g. power source).
This can happen because they are booted at the same time (power returns,
or scheduled dedicated work ends). It can also happen because the daemons
are more or less simultaneously HUPped to reread their configuration files
(block yet another spammer), which is a case I have been noticing locally.
Maybe this doesn't matter much most of the time, but potentially there
could be unnecessarily large surges of network activity. Is there any
easy way to get the queue runs desynchronised? One could use slightly
different -q<time>'s on each host, of course. Would it be worth having
the daemon introduce some pseudo-random fuzz into the interval between
queue runs?
Chris Thompson Cambridge University Computing Service,
Email: cet1@??? New Museums Site, Cambridge CB2 3QG,
Phone: +44 1223 334715 United Kingdom.
