I have been having this issue on and off and can’t really pinpoint what may cause that, but our render farm computers on some files would not finish rendering their buckets but move on to the next ones, resulting in rendering failure.
For example an 8-core machine would start 8 buckets and then keep progressing to another buckets, keeping first 8 highlighted as ‘in progress’; in the end most of the image is covered with unfinished buckets.
When I kill the DR service on the problematic ones, my main workstation will pick up the unfinished buckets and complete the render but I am unable in that case to complete the image using render farm.
Also a note - all works fine in the prepass stages (light cache and irradiance map) - the problem occurs in the final rendering pass only.
It is scene-dependent; Sometimes it works fine. Can’t pinpoint what setting in the scene is common for the failing ones.
I think the problem never happens on rendering passes without GI engine/prepass.
Hello,
I had exactly the same issue. The distributed rendering went fine until i increase the render size to 6000X6000 px. I don’t know if it’s connected because it’s not the first time that I have the problem.
Have you found a solution?
Thank you
Pierre
The previous DR bucket case was connected to the PSD manager plugin. It was missing from some of the render nodes.
If this is not the same case and you have a scene that reproduces the issue, please send us a scene or scenes to test with.