|
|
> >> Client sends > >> arp who-has 192.168.2.2 tell 192.168.2.4 > >> > >> Server receives > >> 65:1c:4a:c8:08:06 null > ff:ff:ff:ff:00:ff sap 08 I (s=3,r=2,C) len=22 > > > As no one else has reported anything about corrupted packets, it's really > hard to tell where the problem is. > > You will have to narrow down the problem. > > Try connecting with another machine. > > Try using an other version of openvpn. > > Try to hook up the client mashine with just a cross-over cable to the > openvpn server to make sure it's not a network problem between the client > and server. > > and so on until you can say exactly under which circumstances the problem > occurs, and when it doesn't. > > If I understood you correct the problem occurs every time you, so it > shouldn't be too hard to narrow it down... Now I can tell a little bit more about the situation, but it doesn't fix the problem. If I use the same configuration on an other machine, it is running fine. Now to my setup: The working environment: Debian woody Celeron 2GHz Realtek The needed/problematic environment: SuSE9.1 pro VIA 1GHz CPU on a EPIA CL-1000 mainboard, which has 2 integrated VIA nics Every time I connect to the SuSE91. machine I will get errors. I also tried to recompile the bins, but without success. The problem is, that I cannot switch the EPIA environment, because it's a productive machine.... The only thing I could determine, is the connection is running if I get following error on both machines: Wed Sep 1 10:08:07 2004 us=427959 Note: Cannot ioctl TUNSETIFF tap0: Invalid argument (errno=22) Wed Sep 1 10:08:07 2004 us=428200 Note: Attempting fallback to kernel 2.2 TUN/TAP interface But because the error occurs only some times I cannot try to reproduce it. Is there any option to force 2.2 TUN/TAP device? Any other hints or suggestions? ____________________________________________ Openvpn-users mailing list Openvpn-users@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/openvpn-users |