http://bugzilla.spamassassin.org/show_bug.cgi?id=4260





------- Additional Comments From [EMAIL PROTECTED]  2005-04-26 21:17 -------
Subject: Re:  rewrite DNS code to use a single socket, event-based model

On Tue, Apr 26, 2005 at 07:23:00PM -0700, [EMAIL PROTECTED] wrote:
> Also, I just ran 3.1 w/out any patches and it took ~3h, so I'm thinking the
> timing comparison from yesterday was a fluke.

Another FYI, I just svn updated to the latest trunk w/ quinlan's rand()
change, same problem after roughly the same time, as expected:

[...]
status:  80% ham: 2674   spam: 17862  date: 2005-04-23   now: 2005-04-26 
11:42:42 PM
uridnsbl: bogus rr: domain=5jb.net, zone=multi.surbl.org.,
rr=dog.ccpatoncejk.biz.      1200    IN      A    200.149.11.62 at
/home/felicity/SA/corpus/../spamassassin-head/masses//../lib/Mail/SpamAssassin/Plugin/URIDNSBL.pm
line 635.
[...]

I'm trying the reinit idea now and seeing how that works.  I made it
counter==orig_counter instead of == 0 since another int isn't really a big
deal.  I'll post tomorrow when I see how it worked out.

I've been thinking about this some more -- why do we care what the
starting ID is?  Why not start at 1, then when we wrap to 0 (16-bit limit)
get a new socket and start anew at 1?  We're not sharing a single socket
between children, so the ID is really only useful to know when responses
from one message are coming in late to the same child.  If we're worried about
another child reusing another's socket, the ID doesn't really help anyway
since the IDs in both children could have been near-similar anyway.





------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

Reply via email to