Computing system for hierarchical task scheduling
AssigneeHuawei Technologies Co., Ltd.
InventorPradeep Jagadeesh
Life of this patent
Estimating the term…
Abstract
A computing system comprises a task scheduler in a top-level, at least two cluster controllers in an intermediate-level, and a plurality of nodes in a bottom level. A task scheduler is configured to receive a task request, obtain an adapted task request from the received task request, based on scheduling information stored in the task scheduler, and forward the adapted task request to one of the at least two cluster controllers. Each cluster controller is configured to receive the adapted task request, obtain an adapted cluster task request from the received adapted task request based on cluster scheduling information stored in the cluster controller, and directly or indirectly forward the adapted cluster task request to a node in the plurality of nodes.
Claims20
A computing system for hierarchical task scheduling, comprising:
- a processor in a top-level layer and configured to:
- store scheduling information;
- receive a task request;
- obtain an adapted task request from the task request based on the scheduling information and the task request by removing information overhead only used for processing at the top-level layer from the task request as determined by the scheduling information;
- and
- forward the adapted task request based on the scheduling information;
- a plurality of nodes in a bottom-level layer, wherein the nodes are coupled to the processor;
- cluster controllers in an intermediate-level layer and coupled to the processor and the nodes, wherein each cluster controller of the cluster controllers is configured to:
- store cluster scheduling information;
- receive, from the processor, the adapted task request;
- obtain an adapted cluster task request from the adapted task request based on the cluster scheduling information and the adapted task request by removing information overhead only used for processing at the cluster controller from the adapted task request as determined by the cluster scheduling information;
- and
- forward the adapted cluster task request to a respective sub-cluster controller in the intermediate-level layer based on the cluster scheduling information;
- and
- sub-cluster controllers in the intermediate-level layer and coupled to the cluster controllers, and wherein each sub-cluster controller of the sub-cluster controllers is configured to:
- store sub-cluster scheduling information;
- receive the adapted cluster task request from a respective cluster controller of the cluster controllers;
- obtain an adapted sub-cluster task request from the adapted cluster task request based on the sub-cluster scheduling information by removing information overhead only used for processing at the sub-cluster controller from the adapted cluster task request as determined by the sub-cluster scheduling information;
- and
- forward the adapted sub-cluster task request to a respective node of the nodes based on the sub-cluster scheduling information.
A hierarchical task scheduling method comprising:
- receiving, by a processor in a top-level layer, a task request;
- obtaining, by the processor, an adapted task request from the task request based on scheduling information stored in the processor and the task request by removing information overhead only used for processing at the top-level layer from the task request as determined by the scheduling information;
- forwarding, by the processor, the adapted task request to a respective cluster controller of cluster controllers in an intermediate-level layer based on the scheduling information;
- receiving, by the respective cluster controller, the adapted task request;
- obtaining, by the respective cluster controller, an adapted cluster task request from the adapted task request based on cluster scheduling information stored in the cluster controllers by removing information overhead only used for processing at the cluster controller from the adapted task request as determined by the cluster scheduling information;
- forwarding, by the respective cluster controller, the adapted cluster task request to a respective sub-cluster controller of sub-cluster controllers in the intermediate-level layer based on the cluster scheduling information;
- receiving, by the respective sub-cluster controller, the adapted cluster task request from the respective cluster controller of the cluster controllers;
- obtaining, by the respective sub-cluster controller, an adapted sub-cluster task request from the adapted cluster task request based on sub-cluster scheduling information and the adapted cluster task request by removing information overhead only used for processing at the sub-cluster controller from the adapted cluster task request as determined by the sub-cluster scheduling information;
- and
- forwarding, by the respective sub-cluster controller, the adapted sub-cluster task request to a respective node of in a plurality of nodes in a bottom-level layer based on the sub-cluster scheduling information.
Dependencies are inferred from the claim text.