> I’m TH and am interested with this > http://www.beowulf.org/pipermail/beowulf/2005-January/011626.html. I’m > currently looking at a solution to launch an object detection app on Host, > with the GUI running on Host and the compute nodes doing all the video > processing and analytics part. I see this solution is similar to my approach > but I couldn’t find any information on how to setup. May I know where to > more info about this? As I googled, Cinelerra is a video processing editor, > which has nothing to do with clustering. Can you guide me on this? > Appreciate your help. >
What is the difference between processing some video and processing some data? With conventional supercomputing the workflow looks like: Transfer data to shared strorage that can be accessed from all nodes. Make a job and submit this to the job scheduler. Wait for the job to be run. Collect the result from a file on the shared storage. Supercomputing is generally not set up for interactive sessions however if your job scheduler always has available slots then the latency to return the data to the app should be minimal. The job file will describe inputs and outputs. Thanks, Andrew > > > Thanks. > > > > FYI, I have tried to send jobs(with graphics) using MPI previously but > failed, seems like MPI only suitable for data processing but not video > processing. > > > > > > > _______________________________________________ > Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf@beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
