AWS Thinkbox Discussion Forums

custom plugin setting: autokill based on machine ram usage

We are implementing this currently, that if nuke/3dsmax is rendering on a machine and the box’s ram usage reaches the physical ram (minus 1g), it kills the render.

With the advent of 64gigs, processes dont crash any more, they just go into swapping mode… bit of a pita, some artist workstations for example are completely locked up in the morning due to heavy jobs, sometimes it just takes 5-10 longer then average to render (due to swapping). Would be nice if deadline had some ‘safety’ controls for this, where we could define a list of plugins (in order of priority) that could be killed when the machine is reaching its ram limit…

So it would first kill nuke, and if its still not enough, kill max etc. (in case multiple slaves are rendering)

Privacy | Site terms | Cookie preferences