AWS Thinkbox Discussion Forums

Moving tasks to faster machines without losing render progress

Hey,

Currently our farm has a variety of different cores on the computer, some task will finish much quicker then others. I was wondering if there is a feature where I could switch out slaves in progress without requesting the entire task. So lets say task 1 finishes on fast computer A, would i be able to swap out task 2 who is currently at 50 percent on slow computer B to continue the render on fast computer A

Thanks

There are very few renderers that allow check-pointing a render to disk so you can continue on a different machine. I believe Maxwell and V-Ray are the two that do here (someone please correct me if I’m wrong there). The other piece is that we would need to support that, which I believe we do for Maxwell in some circumstances, but I’ll need to check on V-Ray.

The short answer is “no”, but can you give me some more details about what you’re moving around to other machines? Also, are these tasks already at one frame each or could you theoretically break out that last task into smaller chunks?

Privacy | Site terms | Cookie preferences