One additional piece of information that I just found out...
When I add the directive on the server-side --push "redirect-gateway" I do not receive this problem. This leads me to believe that the client may have a routing issue? I'm not sure. Hope that helps someone, though.
Date: Mon, 27 Aug 2007 13:03:04 -0400
Subject: [Openvpn-users] OpenVPN Client Flooding Server with UDP Packets 10, 000 per Second for 30 Seconds
I am having a problem with OpenVPN. I have setup ethernet-bridging and am able to connect ok to the server, and have verified that it works (file transfers, mail, etc). However, after about 1 min or so, the connection gets interrupted. I noticed that I wasn't able to browse the network anymore, or check mail, or file transfers were interrupted.
I ran a 'ping vpnserver.mydomain.com -t' to see what was happening. The ping would run successfully at the begining of the VPN connection, but after about 1 min, it would give me 'request timeout' or 'destination not reachable.'
So I put Wireshark on my box to figure out what was going on the line. After looking at it, I was shocked! When the connection was interrupted, there was a flood of UDP packets coming from the client (on ports around 49xxx). These packets would come in approx 10,000 per second for 30 seconds. I do not know what type of packet they are or what they mean. Here is the information that I can give:
Client: Windows Vista
Client Config File:
tls-auth "C:\\Program Files\\OpenVpn\\Keys\\ta.key" 1
ca "C:\\Program Files\\OpenVpn\\Keys\\ca.crt"
cert "C:\\Program Files\\OpenVpn\\Keys\\client.crt"
key "C:\\Program Files\\OpenVpn\\Keys\\client.key"
Server: CentOS 5
Server Config File
tls-auth /etc/openvpn/keys/ta.key 0
server-bridge defaultgw.mydomain.com 255.255.255.0 xxx.xxx.xxx.253 xxx.xxx.xxx.254
push "dhcp-option DNS xxx.xxx.xxx.xxx"
push "dhcp-option DOMAIN mydomain.com"
keepalive 5 120
Thank you for your help in advance...