Priority Wonky

[im_thatoneguy]

I have 2 machines which are rendering on the “next” job even though the current job has unassigned tasks. They are the same groups, pool and priority and the submit time is higher on the later job. But for some reason they really really want to render the later job. Seems like an odd quirk. Any idea why they would skip ahead? They aren’t blacklisted. They aren’t marked as failed. They aren’t anything in particular special.

[jgaudet]

Is there a machine limit on that first job? Are the slaves working on the new job marked as ‘bad’ on that first Job (under Failure Detection)?

[im_thatoneguy]

Nope and Nope.

[MikeOwen]

Do these 2 machines which are rendering the wrong job first, each have a different order of assigned pools or groups to each other? ie: “3d, 2d, etc” & “2d, 3d, etc”