Timeout problems

[r.felber at ek-muc.de]

On Wed, Oct 20, 2010 at 11:39:01PM +0200, r.felber at ek-muc.de wrote:
> On Wed, Oct 20, 2010 at 01:45:15PM -0700, Eric Deis wrote:
> > There is nothing relevant in /var/log/messages, definitely no errors. 
> > Don't have the other logs mentioned.
> > 
> > I included an excerpt previously of what was happening in maillog:
> > Oct 19 21:14:37 MYSERVER postfix/smtpd[5696]: connect from
> > smtp14.icpbounce.com[216.27.93.112]
> > Oct 19 21:16:23 MYSERVER postfix/smtpd[5696]: warning: timeout on
> > 127.0.0.1:12525 while reading input attribute name
> > Oct 19 21:16:23 MYSERVER postfix/smtpd[5696]: warning: problem talking to
> > server 127.0.0.1:12525: Connection timed out
> > Oct 19 21:18:04 MYSERVER postfix/smtpd[5696]: warning: timeout on
> > 127.0.0.1:12525 while reading input attribute name
> > Oct 19 21:18:04 MYSERVER postfix/smtpd[5696]: warning: problem talking to
> > server 127.0.0.1:12525: Connection timed out
> > Oct 19 21:18:04 MYSERVER postfix/smtpd[5696]: NOQUEUE: reject: RCPT from
> > smtp14.icpbounce.com[216.27.93.112]: 451 4.3.5 Server configuration
> > problem; from=<bounces+454705.29871415.356246 at icpbounce.com>
> > to=<gabriela at MYCLIENTDOMAIN.COM> proto=ESMTP helo=<smtp14.icpbounce.com>
> > Oct 19 21:18:04 MYSERVER postfix/smtpd[5696]: disconnect from
> > smtp14.icpbounce.com[216.27.93.112]
> > 
> > 
> > Doing some more research, perhaps my local DNS cache was not setup properly.
> 
> I don't think that was the issue.
> 
> > /etc/resolve.conf was missing the following line:
> > nameserver 127.0.0.1
> > 
> > Only had references to my actual name servers. BIND is running. Did some 
> > dig queries, and looks like the cache is working. First query will take 
> > 100-200ms or less. Second query 1ms or 2ms, indicating cache.
> 
> 
> > I will just wait and see if the errors show up again. Quite random. Once 
> > in three months, then once per week, and then daily. Is there anyway to 
> > test and force the error?
> 
> Once that error occurs frequently one could setup a strace which follows
> forks on the master and its children.
> 
> I see that not as a realistic option.
> 
> I will insert some debug symbols for the listen-stuff in the next
> days. 

Can you please try the version from

http://rz.ek-muc.de/pub/policyd-weight-debug

Warning: it produces much more logging output.


 
-- 
	Robert Felber, PGP: D1B2F2E5