[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

PPTP Issues in Debian Testing/Squeeze



Hello,
 
I am wondering if anyone has come across a similar issue to one I am encountering with Debian Testing/Squeeze.
 
I have been using Debian for many years as a Firewall for my family's computers.
 
I have a ppp interface as the Internet interface and another Ethernet / WiFi interface within my house.
 
Originally, the ppp interface was a dial-up modem, but since the advent of broadband it became a PPTP tunnel forwarded from my ADSL modem.
 
This had been working fine on Debian system I set up many years ago, but recently that PC died.
 
I have built a new system and installed Debian Testing on it (because it should shortly be released as Debian Squeeze and I hoped that if I got Testing working the upgrade to Squeeze should be relatively pain-free).
 
Unfortunately, my network speed appears to be less than 1/10th of what it was previously, even though I believe that I copied all the ppp configuration files over accurately from the old system.
 
I put ppp into debug mode and have been looking at the logs.
 
I get a huge number of lines in my syslog saying:
 
Dec 27 12:51:54 xxxxx pptp[5043]: anon log[decaps_gre:pptp_gre.c:414]: buffering packet 20895 (expecting 20894, lost or reordered)
 
(the packet numbers change).
 
I have tried leaving the MTU unset in the ppp configuration files, and I have also tried setting it to 1416, 1454, 1492 and 1500 (which seem to be common values recommended on the Internet).
 
I am using the Shorewall firewall with the "two-interfaces" configuration.  I have set IP_FORWARDING=On and CLAMPMSS=yes in shorewall.conf.
 
I am uncertain if the issue is with ppp, pptp, Shorewall or something else.
 
I am also uncertain if this could be an issue with the "Testing" distribution as I have never seen anything like this with any of the previous Debian distributions I have used.
 
I would be grateful if anyone has seen this before and could let me know what the issue could be.
 
Regards,
Roland Rosier 		 	   		  

Reply to: