[ previous ] [ next ] [ threads ]
 From:  Manuel Kasper <mk at neon1 dot net>
 To:  "T. Lechat" <m0n0wall at lechat dot org>
 Cc:  "Rijs, Ramon" <ramon at soetec dot nl>, m0n0wall at lists dot m0n0 dot ch
 Subject:  Re: [m0n0wall] ping time outs
 Date:  Thu, 12 Feb 2004 11:05:47 +0100
T. Lechat wrote:

> I have exactly the same problem:
> Here is description :
> - Matt traceroute from outside shows more than 18% of lost packet on WAN
> PPTP m0n0wall interface when traceroute to it dynamic wan ip adress.
 > ...
> -  Each of following things resolve (alone) succesfully the problems:
>    * m0n0wall reboots
>    * ADSL modem reboot ,
>    * Click on 'save' button of M0n0wall Webgui WAN interface page .

Looks like either MPD or the Netgraph ng_pptpgre node is crap - they 
haven't really fixed the Windows XP PPTP VPN problems yet either (the 
patch made it better but didn't completely solve it), despite lots of 
people reporting it on the MPD sourceforge page, and maybe this is even 
related to your PPTP client problems?

Trouble is, the only alternative, poptop, is terribly inefficient since 
it handles all data in userland... :(

Sorry, but I don't consider this a m0n0wall problem, but rather one in 
FreeBSD and/or MPD that I probably can't fix (OK, I thought that about 
three times before when there were bugs in ipfilter, as well as the 
FreeBSD kernel (dummynet+ipf) and still came up with a solution after 
trying hard for long enough... :)

- Manuel