Re: Spamass-milter stalls

[Andrew Billinghurst]

On 6 Sep 2003 at 11:54, Dan Nelson wrote:

> In the last episode (Sep 07), Andrew Billinghurst said:
> [bad linewrapping manually corrected)
> > On the test server, sendmail is not consulting RBLs.  It is not even 
> > getting a full outside feed.
> > 
> > First error message is
> > 
> > Sep  3 07:09:37 testserver sendmail[11899]: h83D5amB011899: Milter 
> > (spamassassin): timeout before data read
> > Sep  3 07:09:37 testserver sendmail[11899]: h83D5amB011899: Milter 
> > (spamassassin): to error state
> > Sep  3 07:09:37 testserver sendmail[11920]: h83D5amB011899: 
> > to=<address@hidden>, delay=00:04:01, xdelay=00:00:00, mailer=local, 
> > pri=31993, dsn=2.0.0, stat=Sent
> > 
> > It later changes to
> > 
> > Sep  3 15:59:56 testserver sendmail[15919]: h83LiumB015919: Milter 
> > (spamassassin): error connecting to filter: Connection timed out with 
> > /var/run/spamass.sock
> > Sep  3 15:59:56 testserver sendmail[15919]: h83LiumB015919: Milter 
> > (spamassassin): to error state
> > Sep  3 15:59:56 testserver sendmail[15919]: h83LiumB015919: 
> > internal.rootsweb.com [192.168.16.42] did not issue MAIL/EXPN/VRFY/ETRN 
> > during connection to MTA
> > 
> > Which to me says not a lot in terms of problem solving.
> 
> Has it crashed?  If "ps ax | grep spamass-milter" returns less than
> three processes, then it segfaulted and Linux's threads library failed
> to kill the remaining processes.

Crashed and the two processes have gone to electron heaven, left with 
one.  I will presume that means that it segfaulted.  Sendmail 
basically runs up to the maximum no. of processes and stops.  The 
process 
> 
> What version of spamass-milter are you running?  If you're using a CVS
> version, what's the $Id$ line at the very top of spamass-milter.cpp?
> 

Using the last stable version listed, Version 0.2.0. 

$Id: spamass-milter.cpp,v 1.52 2003/06/26 15:10:44 dnelson Exp $

Regards, Andrew