After submiting job, i’m analysing ram usage to control the maximum concurrent task limit per machine. This value varies from project to project due to n factors. So, after i check the maximum ram usage, i change the values for the servers and I thought it would start automatically to render queued taks. But it doesn’t… can I force this? how? Can this be done automatically? How?
I am not quite sure I understand the question. Are you asking how to force the changes to take effect? You might need to have the slave instance restart for that, at worst, or at best, it should be automatic.
Imagine this scenario: 1 - Submit job with 40 partitions krakatoa based on pflow and fumefx to drive particle motion. I have 4 licenses and i start with 2 concurrent task limit (so, we will have 8 queued tasks). Then i check ram usage and decide to change this limit per machine. I thought that when I change this (imagine for 4 concurrent task limit per server), deadline would send instructions automatically to this 4 servers and 8 more tasks would start… it’s not.
So, if i restart slave, will this mean that i will loose all the frames rendered until i make that instruction?
We confirmed that the slave will only recognize that its concurrent task limit override has changed when it starts a NEW job. We will fix this for Deadline 6.2, which is currently in beta: thinkboxsoftware.com/news/20 … mx-20.html
You can find instructions in the link above if you’re interested in joining.