http://bugzilla.spamassassin.org/show_bug.cgi?id=3077





------- Additional Comments From [EMAIL PROTECTED]  2004-03-08 08:32 -------
Subject: RE:  spamassassin -d is too damn slow




>
>  Additional Comments From [EMAIL PROTECTED]  2004-03-07 15:49
>
> I'd be interested to see how it does right now. :)

It looks good. Here's the scoop. Running as 'root' to avoid potential NFS
mounts of home dir., specifying -p /dev/null to avoid issues relating to
user prefs.:


# time spamassassin -d --mbox -p /dev/null < spam-1000-msgs.mbox >
spam-1000-clean.mbox
Created user preferences file: /dev/null
6.490u 0.420s 0:08.25 83.7%     0+0k 0+0io 593pf+0w


That's 102 messages per. sec. on a 2.4G P4 Xeon running RH 9.

Compare this to the results in this note:
http://bugzilla.spamassassin.org/attachment.cgi?id=1795&action=view

which yielded the following performance given the same dataset, under
similar circumstances:

# time formail -s spamassassin -L -d -C ./empty -p ./prefs <
spam-1000-msgs.mbox > spam-1000-clean-2.mbox
422.000u 37.420s 7:40.09 99.8%  0+0k 0+0io 574120pf+0w

which limped along at 2.1 msgs/sec.

Thus the new --mbox method is 47 times faster than the old one-at-a-time
method, and 2x as fast as the formail/procmail/sed method (which doesn't do
everything that SA -d does.)

I'd say that's a noticeable improvement. Nice work Theo, et al.

   - Gary






------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

Reply via email to