Centaurus Edge should support flexible workload scheduling, and this could be achieved by adopting ideas from the global scheduling project including resource info gathering, scheduling algorithm, and scheduler pool cooperation. To name a few of the benefits:
Smart workload placement where end user do not need to specify a specific cluster but instead the resource requirement along with other labels/affinities (e.g. "I need to run 5 replicas, each requires xxx cpu and yyy memory in region/area zzz")
Centaurus Edge should support flexible workload scheduling, and this could be achieved by adopting ideas from the global scheduling project including resource info gathering, scheduling algorithm, and scheduler pool cooperation. To name a few of the benefits: