Remote use of XDMCP
I currently have XDMCP running well on my local network, in that computers on my home network can use it to access other computers. However, when I try to connect via WiFi & VPN, one computer that works well on the local lan, cannot connect. I've used ethereal to see what's happening. In the list below, the letters A & B indicate the source of the packets. "A" (10.x.x.x address) is a notebood connected via WiFi & VPN and "B" (192.168.x.x) is my main desktop system. Both systems are SUSE 10.0. Ethereal, running on both the notebook and desktop shows the following. A XDMCD Query B XDMCP Willing A XDMCP Request B XDMCP Accept A XDMCP Manage (repeats 7 times) B XDMCP Failed. So, it's obvious the two systems are communicating, but unable to run an X session. Is there something that prevents a routed connection from working? Perhaps an address range issue? Other ideas? tnx jk
On Sunday 19 March 2006 17:46, James Knott wrote:
So, it's obvious the two systems are communicating, but unable to run an X session. Is there something that prevents a routed connection from working? Perhaps an address range issue? Other ideas?
One problem I saw when I tried to use xdmcp on a machine with two network cards is that X will try to use the first IP address it find on the local machine. So it may be X isn't going over the VPN. You should be able to see that in the ethereal trace though. If you try X -from <VPN interface IP address> -query <server IP> does it work then? -- Certified: Yes. Certifiable: of course! jabber ID: anders@rydsbo.net
Anders Johansson wrote:
On Sunday 19 March 2006 17:46, James Knott wrote:
So, it's obvious the two systems are communicating, but unable to run an X session. Is there something that prevents a routed connection from working? Perhaps an address range issue? Other ideas?
One problem I saw when I tried to use xdmcp on a machine with two network cards is that X will try to use the first IP address it find on the local machine. So it may be X isn't going over the VPN. You should be able to see that in the ethereal trace though.
If you try
X -from <VPN interface IP address> -query <server IP>
does it work then?
The 10.x.x.x address I mentioned is the VPN interface address. The WiFi address is 192.168.2.166.
On Sunday 19 March 2006 20:02, James Knott wrote:
The 10.x.x.x address I mentioned is the VPN interface address. The WiFi address is 192.168.2.166.
Thanks for the info, but you didn't answer my question. Does it work if you try starting X the way I suggested? -- Certified: Yes. Certifiable: of course! jabber ID: anders@rydsbo.net
Anders Johansson wrote:
On Sunday 19 March 2006 20:02, James Knott wrote:
The 10.x.x.x address I mentioned is the VPN interface address. The WiFi address is 192.168.2.166.
Thanks for the info, but you didn't answer my question. Does it work if you try starting X the way I suggested?
Yes it does work from the command line. I wonder if there's a way around that, for the SUSE remote login?
On Sunday 19 March 2006 20:28, James Knott wrote:
Yes it does work from the command line. I wonder if there's a way around that, for the SUSE remote login?
Short of hacking kdm to add the -from as an option, I suspect the only way is to force the VPN interface to be the first available, so X will bind to it. At the moment I'm not entirely sure how to do that though -- Certified: Yes. Certifiable: of course! jabber ID: anders@rydsbo.net
Anders Johansson wrote:
On Sunday 19 March 2006 20:28, James Knott wrote:
Yes it does work from the command line. I wonder if there's a way around that, for the SUSE remote login?
Short of hacking kdm to add the -from as an option, I suspect the only way is to force the VPN interface to be the first available, so X will bind to it. At the moment I'm not entirely sure how to do that though
That might be a bit tough, as the VPN requires some other network device to carry it. I suspect a better method would be to watch where the XDMCP ACCEPT comes from.
On Sunday 19 March 2006 20:43, James Knott wrote:
That might be a bit tough, as the VPN requires some other network device to carry it.
Shouldn't matter, since the big thing is the order in which they are presented when an application requests an interface. I don't think that would affect the VPN functionality
I suspect a better method would be to watch where the XDMCP ACCEPT comes from.
This would require changes to either kdm or X (and I know which one I'd rather dig into, I've seen the sources to both :) -- Certified: Yes. Certifiable: of course! jabber ID: anders@rydsbo.net
At 06:52 AM 20/03/2006, Anders Johansson wrote:
On Sunday 19 March 2006 20:43, James Knott wrote:
That might be a bit tough, as the VPN requires some other network device to carry it.
Shouldn't matter, since the big thing is the order in which they are presented when an application requests an interface. I don't think that would affect the VPN functionality
I suspect a better method would be to watch where the XDMCP ACCEPT comes from.
This would require changes to either kdm or X (and I know which one I'd rather dig into, I've seen the sources to both :)
what about asking Matthias if anything can be dine from his end? 2bits scsijon
participants (3)
-
Anders Johansson -
James Knott -
scsijon