We have a job that has a weight of 2300 (the lowest in the queue), but it keeps getting new machines assigned to it.
Is there a way to troubleshoot why the machines decide to pick that job?
Scheduling is set to pool / weight / balanced. Time is not a factor.
There are other jobs that those machines could be rendering, with a higher weight.
It could be that the limit groups have been filled up for those other jobs. With the constant self updating, its hard to get a factual snapshot state of whats happening 
Yeah, limit groups or machine limits could definitely be playing a role here. If you have slave verbose logging enabled, you can check out the slave log to see which jobs it’s skipping over due to maximized limits.
Had this problem again. We have about 230 queued jobs, but machines are picking up a low weighted job continually.
Attached is the job data files.
job.zip (10.4 KB)
That particular job’s machine limit is maxed out, but from the email we got from you guys, it sounds like your issues are boiling down to the database being overwhelmed by the number of Monitors connecting to it. We’re currently discussing ways to deal with this.
Cheers,
Ryan
This was a weight job, and thats my point. It should not have been rendering, yet its machine limit was maxed out, while other jobs with higher weights were not doing anything.