Index: [Article Count Order] [Thread]

Date:  Tue, 16 May 2006 16:16:09 +0100
From:  Simon Dick <simon.dick (at mark) advantage-interactive.com>
Subject:  [coba-e:05213] Re: The ongoing QPOPPER Debate
To:  coba-e (at mark) bluequartz.org
Message-Id:  <1147792569.77214.2.camel (at mark) sdwork.office.ai270.net>
In-Reply-To:  <101476125.20060516164959 (at mark) gmx.net>
References:  <076801c678e7$c67fa370$0301a8c0 (at mark) Jerrycp>	 <101476125.20060516164959 (at mark) gmx.net>
X-Mail-Count: 05213

A lot of times I've seen issues with qpopper I've also seen errors with
the pam layer on the 550 os, having not got any bq boxes in production
use I can't say if that's still the problem or not

On Tue, 2006-05-16 at 16:49 +0200, Chris Hemsing wrote:
> Hello Gerald,
> 
> to my knowledge, this issue has not yet been solved.
> 
> I totally agree, that raq4 with an ancient 450MHz AMD K6 does not have these problems
> with hundreds of users!
> Thus it is not really a qpopper itself issue!
> 
> As it seems nobody has ever done a thourough analysis.
> Like
> 1) what is the situation on the network side (trace the traffic that
> leads to the situation)
> 2) if qpopper gets stuck, has anybody attached a debugger and looked
> at what the situation in qpopper is?
> 3) qpopper compiled with debugging information
> 4) have qpopper do extensive logging/debugging output?
> 5) differentiate whether there is also an xinetd problem (or rather adjustment
> problem). The default maximum number of connections per second xinetd
> will accept is 50. If a company does a burst in parallel for all its
> mailaccounts...and this happens in parallel with some companies....
> 
> Instead of doing all the work to move to a different pop3 solution
> (and possibly end up with the same situation as with qpopper) one
> should thouroughly  analyse the problem. I would like to assisit, but
> I don't have the problem.
> 
> If someone can give me ssh access to a machine, where the problems
> regularly show up, I will volunteer to try to find the reason.
> 
> Chris
> 
> Tuesday, May 16, 2006, 2:53:47 PM, you wrote:
> 
> GW> Ken Marcus wrote
> >> I changed this on 5 of my servers and  it did not have any 
> >> ill effects that 
> >> I can detect.
> >> 
> >> I have not had any problems with it in any case, but if the 
> >> reverse on the 
> >> FTP slows it up so much, then it stands to reason that if the 
> >> qpopper is 
> >> doing a reverse lookup that it would slow it down immensly.
> >> 
> >> On raq4 servers in the past I have had problems with 
> >> customers that use 
> >> something like popbeamer and overwhelm the qpopper. Or whole 
> >> offices with 
> >> time synchronized computers all checking their email every minute.
> 
> GW> We have customers that are still having issues with qpopper.
> GW> Has there been a fix for this problem.
> 
> GW> One customer has a RaQ4 that can handle the pop3 connections,
> GW> but a MSI server 3.0GHz Pentium4 2GB RAM can not!
> 
> GW> Gerald
> 
>