> From: Gary Funck [mailto:[EMAIL PROTECTED]
> Sent: Monday, February 23, 2004 4:38 PM
>
[...]
> Bart Schaefer recommended that I try 'pperl':
>
> % time formail -s pperl `which spamassassin` -L -d -C ./empty -p ./prefs <
> spam-1000-msgs.mbox > spam-1000-clean-2.mbox
> User=4.540 System=2.220 Wall=1:14.88
>
> Which is quite a bit better -- 6.4x times faster than our first,
> naive try.
> This brings the processing rate up to 13.5 messages per second, which is
> 1/8-th
> of the formail/procmail/sed solution, but is much improved.
I should add that this test was run on a 2 processor Xeon, which
has to physical cpus + 2 virtual (hyper-threaded) cpus. Thus, if
pperl works by letting each sub-process run (it forks 5 by default)
then it possible that a good deal of the speed up was derived from
increased parallelism.
