# Indirect rendering when using GLX from remote machine

## post by mikeramsey on Jan 20, 2021

Hi there:

I’m remoting into our machine (that contains 4 nvidia cards) using ssh -X machine.

I’m trying to get _**glxinfo**_ working as an initial test for some further engine development and I’m getting the following error text after issuing _**glxinfo**_:

_name of display: localhost:10.0_

_X Error of failed request: BadValue (integer parameter out of range for operation)_

_Major opcode of failed request: 151 (GLX)_

_Minor opcode of failed request: 24 (X_GLXCreateNewContext)_

_Value in failed request: 0x0_

_Serial number of failed request: 41_

_Current serial number in output stream: 42_

After looking on the internet it appears to be that the support for X11 servers OpenGL indirection is _not_ enabled on the lambda by default.

I’ve tried adding the below section to /usr/share/X11/xorg.conf.d/99-allow-indirect-glx.conf

_Section “ServerFlags”_

_ Option “AllowIndirectGLX” “on”_

_ Option “IndirectGLX” “on”_

_ EndSection_

typically that would be in /etc/X11 but I’m not seeing that file there.

So the two questions I have for support are:

1.) Is this truly an issue due to indirect rendering when you use GLX from a remote machine?

2.) And if so, how can I enable indirect rendering on our Lambda machine?

Thanks,

Mike Ramsey

## post by mikeramsey on Jan 21, 2021

Figured it out, you need to modify 10-nvidia.conf. Works like a charm…sorta, at least for simple apps. Old drivers on the lambda hinders engine work.