Simultaneous scheduling of processes and offloading computation on many-core coprocessors

摘要

Methods and systems for scheduling jobs to manycore nodes in a cluster include selecting a job to run according to the job's wait time and the job's expected execution time; sending job requirements to all nodes in a cluster, where each node includes a manycore processor; determining at each node whether said node has sufficient resources to ever satisfy the job requirements and, if no node has sufficient resources, deleting the job; creating a list of nodes that have sufficient free resources at a present time to satisfy the job requirements; and assigning the job to a node, based on a difference between an expected execution time and associated confidence value for each node and a hypothetical fastest execution time and associated hypothetical maximum confidence value.

主权项

1. A method for scheduling jobs to manycore nodes in a cluster, comprising:
selecting a job to run according to the job's wait time and the job's expected execution time; sending job requirements to all nodes in a cluster, wherein each node comprises a manycore processor; determining at each node whether said node has sufficient resources to ever satisfy the job requirements and, if no node has sufficient resources, deleting the job; creating a list of nodes that have sufficient free resources at a present time to satisfy the job requirements; and assigning the job to a node using a processor, based on a difference between an expected execution time and associated confidence value for each node and a hypothetical fastest execution time and associated hypothetical maximum confidence value;wherein selecting a job to run comprises ordering a list of pending jobs according to a priority based on a difference between a time each job has waited and a weighted expected execution time for each respective job; andwherein the priority Pi of each job i is calculated according to Pi=wi−Aei, where wi is the amount of time the job i has been waiting for scheduling, ei is the expected execution time for the job i, and A is a configuration parameter.