I haven't been able to find good info about this on these forums or elsewhere.
I use OpenVPN with certificates to connect to my home network. The connection is tunneled so the remote machine gets an address from my home network on its VPN interface. This all works fine from the Windows box on my desk. The VPN isn't configured to take over... access to the internet and other work resources is normal and remote desktop, FTP, or whatever to machines on the home subnet goes over the VPN.
So I installed Hardy on a laptop and brought it in to test. I use the command line with sudo to start OpenVPN and specify the config file, keys, and certificates. The tunnel comes up with no error messages. Here's where the fun starts.
- The tap0 interface won't request an address automatically (it does on windows)
- running 'sudo dhclient tap0' manually does the trick for that. I get an address from home.
At this point, I can't get to anywhere on the internet, at work, or at home. No pings work, nothing. ifconfig shows good addresses on eth0 and tap0.
For anything to work, i have to run 'sudo dhclient eth0'. The DHCP server responds with the same IP I've had all along. Magically, everything works. I can get to the internet, stuff at work and home.
Why do I have to reset DHCP on eth0 before anything works?
Note: running 'ip route' shows both subnets with their proper default gateways. However, tap0 is the first default listed when I make it grab an address. When I reset DHCP on eth0, it then becomes the first default listed. Does the order in which these are listed matter?
I use OpenVPN with certificates to connect to my home network. The connection is tunneled so the remote machine gets an address from my home network on its VPN interface. This all works fine from the Windows box on my desk. The VPN isn't configured to take over... access to the internet and other work resources is normal and remote desktop, FTP, or whatever to machines on the home subnet goes over the VPN.
So I installed Hardy on a laptop and brought it in to test. I use the command line with sudo to start OpenVPN and specify the config file, keys, and certificates. The tunnel comes up with no error messages. Here's where the fun starts.
- The tap0 interface won't request an address automatically (it does on windows)
- running 'sudo dhclient tap0' manually does the trick for that. I get an address from home.
At this point, I can't get to anywhere on the internet, at work, or at home. No pings work, nothing. ifconfig shows good addresses on eth0 and tap0.
For anything to work, i have to run 'sudo dhclient eth0'. The DHCP server responds with the same IP I've had all along. Magically, everything works. I can get to the internet, stuff at work and home.
Why do I have to reset DHCP on eth0 before anything works?
Note: running 'ip route' shows both subnets with their proper default gateways. However, tap0 is the first default listed when I make it grab an address. When I reset DHCP on eth0, it then becomes the first default listed. Does the order in which these are listed matter?
Comment