Hi all,
I am a vfx artist and know little about networking and IT. I am doing some research into optimizing a rendering system. The scenario is having 25+ dedicated powerful rendering nodes and 800+ computers that can be used as nodes when not in use. Jobs are sent out by a server running the farm software and another server is acting as storage. This is a Linux environment. At the moment, When someone logs into one of the computers that is rendering, the frame is cancelled and resubmitted to another machine by the server. I was wondering if we used virtual machines, maybe the process can be paused, moved to another node, them continued. This way frames are not dropped and render times based on lab activity are reduced.
As I said I know only a bit about networking or virtual machines. I am looking into this for fun and I am not sure if anything will really come of this. Any ideas on were I can find resources to help me learn more about what I need to know? Also what impacts would it have on a network?
I am a vfx artist and know little about networking and IT. I am doing some research into optimizing a rendering system. The scenario is having 25+ dedicated powerful rendering nodes and 800+ computers that can be used as nodes when not in use. Jobs are sent out by a server running the farm software and another server is acting as storage. This is a Linux environment. At the moment, When someone logs into one of the computers that is rendering, the frame is cancelled and resubmitted to another machine by the server. I was wondering if we used virtual machines, maybe the process can be paused, moved to another node, them continued. This way frames are not dropped and render times based on lab activity are reduced.
As I said I know only a bit about networking or virtual machines. I am looking into this for fun and I am not sure if anything will really come of this. Any ideas on were I can find resources to help me learn more about what I need to know? Also what impacts would it have on a network?
Last edited:
