Welcome Nick, and thanks for introducing yourself. It helps a lot to have a bit more context on the person!
On Wed, Jan 28, 2009 at 10:36 PM, NickSmith (via Nabble) <[email protected]> wrote: > I have found much of the help here very useful, till I got to where I needed > to start making config files. I am no coder, and I am finding the syntax > hard to grasp and haven't found anywhere that explains it well. You are looking for the non-existing User Guide, unfortunately. The syntax is on the website (search file format), but that only helps you if you no the meaning already. > I guess what I would like to know is... which of the config examples would > be a good one to start from to get say... EqPly running across all 6 > screens? hactar.eqc > Is that possible? Yes. You can also have a look at the config tool. It can generate some of the basic configuration files, among them display wall configurations (eq './build/Darwin/bin/configTool -n 6 -m Wall -C 3') > Another thing I am having a hard time finding > info on is how to get EQ to fire on the other 5 nodes. For example on the > linux side, I assume it SSH connects to the other nodes and tells them to > run what they need to run in order to display what they are gonna display. > about 75% of what I read on the website looks like a foreign language to me > :) I could go into how I (a non VIS expert and a non-coder) got put on this > project, but that is beside the point, and I doubt ya'll have a day to read > about that one :) I feel your pain, and hopefully you'll find help here. The issue on our end is one of time, and we try to document as much as possible. But unfortunately, we come from the other end towards you. :) So - first to expand on the config file format: In the first section you describe your resources - nodes (Cluster machines), pipes (graphics cards), windows (on/offscreen) and channels (2D render area in a window). In the second section you describe your rendering setup, i.e., how the channels are used and combined to produce renderings (compounds). The first section in a display wall case is straight-forward: you create one node and pipe for each machine and GPU, and one full-screen window with one full-screen channel for each GPU/display. In the second section, you use each of these channels in a compound, and describe the physical size of the display in a wall{} description. Furthermore, you group them all in another compound in order to use a swapbarrier{} for swap synchronization (all display show a new frame at the same time). That's all! Now to launching node processes: On Linux, I recommend using auto-launching with ssh: - make sure you can log into all nodes without a password from the node where you launch the application. - use the same directory for Equalizer on all nodes, e.g., by using NFS - use the node hostnames in the connection{} section of the appropriate nodes. -- some Linux setups resolve your 'hostname' to a local IP, in which case you have to use IP addresses or fix your setup. - Launch eqServer <yourconfigfile> - Launch eqPly On Windows, I recommend manually starting the render clients. Getting ssh set up correctly ain't easy. You can read more about this here: http://www.equalizergraphics.com/documents/design/residentNodes.html HTH, Stefan. -- View this message in context: http://n2.nabble.com/Newbie-here%2C-be-nice-please-%3A%29-tp2235067p2237609.html Sent from the Equalizer - Parallel Rendering mailing list archive at Nabble.com.
_______________________________________________ eq-dev mailing list [email protected] http://www.equalizergraphics.com/cgi-bin/mailman/listinfo/eq-dev http://www.equalizergraphics.com
