Every so often I get nudged to check into a message stuck on one of our
inbound MXes.
So far, every one of them has been spam, but a few have cause some odd
behaviour with spamc/spamd.
Here's one pretty much guaranteed to peg a CPU core for ~130 seconds (or
more):
http://pastebin.com/2ssy2YEk
Is there a way to determine which specific rule is likely to be causing
the excessive scan times? I've already tried flipping rulesets around
and adding some of the timing header report bits; the slowdowns so far
have always happened with the stock rules, and the timing report doesn't
break down nearly far enough.
-kgd
