[Openvpn-users] Bad LZO decompression header byte: 0

[Micha S. <mi...@ar...>]

Hello All:
We've been using OpenVPN for a few years now, quite successfully. I'm 
now on version 2.0.7 on the CentOS 4 server (firewall), and clients are 
using mostly WinXP, some Win2K and some Linux (OVPN versions 2.0.3 - 
2.0.7)  On one new, remote computer - running WinXP - I can't get the 
tunnel to send traffic. It seems to start OK, but as soon as I try to 
connect to anything beyond the FW/OpenVPN server, this error appears, 
and a few seconds later a ping restart:

Fri Jul 28 17:27:07 2006 us=476356 Initialization Sequence Completed
Fri Jul 28 17:27:13 2006 us=367391 Bad LZO decompression header byte: 0
Fri Jul 28 17:27:24 2006 us=129201 Bad LZO decompression header byte: 0
Fri Jul 28 17:27:33 2006 us=641356 Bad LZO decompression header byte: 0
Fri Jul 28 17:27:43 2006 us=389408 Bad LZO decompression header byte: 0
Fri Jul 28 17:27:53 2006 us=671409 Bad LZO decompression header byte: 0
Fri Jul 28 17:28:03 2006 us=790493 [paran.arava.co.il] Inactivity 
timeout (--ping-restart), restarting
Fri Jul 28 17:28:03 2006 us=790927 TCP/UDP: Closing socket
...

The only thing different about this new workstation is a USB wireless 
network adapter. (Although some of the  other remote users have laptops 
with builtin wireless, and they use the VPN successfully.) Another 
point: I'm using mssfix=1420 and fragment=1420 in all configs, as well 
as comp-lzo, of course.

Where can I search to find out what's wrong here?

Thanks,
Micha