Job and Task Scheduling and Resource Allocation on Parallel/Distributed machines by Supervisor

Scheduling parallel jobs onto machines is one of the most important problems to meet any performance objectives in parallel and distributed systems. The objective may be to minimize execution time, but one may want to meet other objectives at the same time, such as good resource utilization, energy-efficiency, income generation, etc. As the problem is NP-complete in general, there is lots of work on heuristics and sub-optimal solutions. The target environments could vary: they could be distributed environments, clouds, multicore machines, GPUs, etc. These different environments change the parameters that lead to good solutions. There is already significant expertise on various forms of DAG scheduling (particularly influenced by workflow scheduling) problems and I am happy to discuss and shape a project related to any type of resource allocation, optimization criterion, and platform (multicore, manycore, cloud, etc).

For past work please check relevant publications from http://www.cs.man.ac.uk/~rizos/publications.html

Keywords: Parallelism, Scheduling, Workflow, DAG