Re: [VirtualGL-Users] How to start a Xserver display by ssh ? and why port 4242 ?

[DRC <dco...@us...>]

On 3/18/13 3:44 PM, First Last wrote:
> it's working!
> I don't understand why nomachine is working... because by default in
>
>     /etc/ssh/sshd_config
>
> forwarding are disabled...
> so I enabled these,
>
>     AllowAgentForwarding yes
>     AllowTcpForwarding yes
>     X11Forwarding yes

NX doesn't rely on the X11 forwarding capability in sshd.  It has its 
own method of compressing and then tunneling the X11 command stream, and 
architecturally, it is actually a combination of an X11 proxy (which 
compresses the X11 stream) and a client-side X server (which takes care 
of displaying the decompressed X11 stream.)

 From Section 6.3 of the VirtualGL User's Guide:

"The application server’s SSH daemon should have the X11Forwarding 
option enabled and the UseLogin option disabled.  This is configured in 
sshd_config, which is usually located under /etc/ssh."


> not sure if all are necessaries, but after that I can
>
>     /opt/VirtualGL/bin/vglconnect user@SERVER
>
> without error. Then (server),
>
>     /opt/VirtualGL/bin/vglrun etracer
>
> (without the "-d" ) option works! I started to try with "-d :0" but it
> looks like rendering in the server itself but show no window.

The default value of VGL_DISPLAY (which is what you're setting with the 
vglrun -d switch) is :0.0.  Not sure why setting it to :0 wouldn't work, 
but it could be a factor of using a headless display.


> I have a small question... not related to virtualgl itself... how can
> get the list of displays ? because in nomachine if I type
>
>     echo $DISPLAY
>
> I get,
>
>     :1000

Not sure what you mean by the "list of displays".