Description
In a cluster with limited resources where production jobs take scheduling priority and/or preempt non-prod jobs, it will be useful to track median times to assigned/running for non-prod jobs. This will help estimating the performance impact in the event of a (un-)planned capacity removal.