Method, device and computer program product for managing dedicated processing resources

Patent metadata

Field	Value
Publication number	US-11663050-B2
Application number	US-201916384209-A
Country	US
Kind code	B2
Filing date	Apr 15, 2019
Priority date	Apr 20, 2018
Publication date	May 30, 2023
Grant date	May 30, 2023

Abstract

A resource management method comprises: in response to receiving, from an application operating on a client, a resource allocation request indicating an amount of dedicated processing resources required by the application, acquiring a mapping between a group of physical dedicated processing resources provided by a group of servers and a group of logical dedicated processing resources, the group of physical dedicated processing resources being divided into the group of logical dedicated processing resources; determining allocation statuses of the group of logical dedicated processing resources; determining, based at least on the mapping and the allocation statuses, a first amount of logical dedicated processing resources to be allocated to the application from the group of logical dedicated processing resources; and indicating the first amount of logical dedicated processing resources to the application, to allow the application to utilize physical dedicated processing resources provided by at least one of the group of servers.

First claim

What is claimed is:

1. A method, comprising:
logically partitioning, by a controller, a group of physical dedicated processing resources, which reside on a group of servers, into a group of logical dedicated processing resources, wherein at least one physical dedicated processing resource is logically partitioned into multiple logical dedicated processing resources to enable concurrent use of the at least one physical dedicated processing resource by a plurality of applications, and wherein at least one physical dedicated processing resource of the group of physical dedicated processing resources comprises a graphics processing unit (GPU) device comprising GPU compute and storage resources that are logically partitioned into a plurality of logical GPU compute and storage resources;
receiving, by the controller, a resource allocation request from a client driver operating on a client device, the resource allocation request indicating an amount of dedicated processing resources for allocating to an application executing on the client device;
in response to the resource allocation request, the controller performing a process comprising:
acquiring a mapping between the group of physical dedicated processing resources and the group of logical dedicated processing resources, wherein the mapping comprises at least one entry which maps at least one logical dedicated processing resource to at least one physical dedicated processing resource and which identifies a server of the group of servers on which the at least one physical dedicated processing resource resides;
determining allocation statuses of the group of logical dedicated processing resources;
determining, based at least on the mapping and the allocation statuses, a first amount of logical dedicated processing resources to be allocated to the application from the group of logical dedicated processing resources, the first amount of logical dedicated processing resources meeting the requested amount of dedicated processing resources;
indicating the first amount of logical dedicated processing resources to the client driver, to allow the client driver to communicate with at least one server of the group of servers to establish at least one data path between the client device and the at least one server of the group of servers and enable the application to utilize physical dedicated processing resources provided by the at least one server of the group of servers, the physical dedicated processing resources corresponding to the first amount of logical dedicated processing resources; and

in response to receiving, by the controller, notification of completion of execution of the application and release of the first amount of logical dedicated processing resources, the controller updating the allocation statuses of the group of logical dedicated processing resources to indicate availability of the first amount of logical dedicated processing resources;
wherein the controller comprises at least one processing device comprising a processor coupled to a memory.

2. The method according to claim 1 , wherein the allocation statuses indicate a second amount of logical dedicated processing resources available for allocation in the group of logical dedicated processing resources, and determining the first amount of logical dedicated processing resources comprises:
determining, from the second amount of logical dedicated processing resources, the first amount of logical dedicated processing resources meeting the amount of dedicated processing resources.

3. The method according to claim 1 , wherein determining the first amount of logical dedicated processing resources comprises:
determining the first amount of logical dedicated processing resources, such that the first amount of logical dedicated processing resources include a plurality of logical dedicated processing resources mapped to a same physical dedicated processing resource in the group of physical dedicated processing resources.

4. The method of claim 1 , wherein determining the first amount of logical dedicated processing resources comprises:
determining the first amount of logical dedicated processing resources, such that the first amount of logical dedicated processing resources include a plurality of logical dedicated processing resources provided by a same server in the group of servers.

5. The method of claim 1 , wherein determining the first amount of logical dedicated processing resources comprises:
determining the first amount of logical dedicated processing resources, such that the first amount of logical dedicated processing resources is uniformly mapped to the group of physical dedicated processing resources.

6. The method of claim 1 , wherein determining the first amount of logical dedicated processing resources comprises:
determining an estimate about execution efficiencies of the application over different physical dedicated processing resources; and
determining, based on the estimate, the first amount of logical dedicated processing resources to be allocated to the application.

7. The method of claim 1 , wherein indicating the first amount of logical dedicated processing resources to the application comprises:
determining whether the first amount of logical dedicated processing resources include a plurality of logical dedicated processing resources mapped to a same physical dedicated processing resource;
in response to determining that the first amount of logical dedicated processing resources include a plurality of logical dedicated processing resources mapped to a same physical dedicated processing resource, aggregating the plurality of logical dedicated processing resources into a single logical dedicated processing resource;
indicating the first amount of logical dedicated processing resources by indicating the single logical dedicated processing resource to the application.

8. The method of claim 1 , further comprising:
in response to the first amount of logical dedicated processing resources being indicated to the application, updating respective allocation statuses of the first amount of logical dedicated processing resources.

9. The method of claim 1 , further comprising the controller generating the mapping and storing the mapping in a storage device that is accessible by the controller.

10. A device, comprising:
at least one processing unit;
at least one memory coupled to the at least one processing unit and storing instructions executed by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the device to perform actions for managing resources, the actions comprising:
logically partitioning a group of physical dedicated processing resources, which reside on a group of servers, into a group of logical dedicated processing resources, wherein at least one physical dedicated processing resource is logically partitioned into multiple logical dedicated processing resources to enable concurrent use of the at least one physical dedicated processing resource by a plurality of applications, and wherein at least one physical dedicated processing resource of the group of physical dedicated processing resources comprises a graphics processing unit (GPU) device comprising GPU compute and storage resources that are logically partitioned into a plurality of logical GPU compute and storage resources;
receiving a resource allocation request from a client driver operating on a client device, the resource allocation request indicating an amount of dedicated processing resources for allocating to an application executing on the client device;
in response to the resource allocation request, performing a process which comprises:
acquiring a m

Assignees

• Emc Ip Holding Co Llc

Inventors

• Peng Layne Lin
• Zhao Junping
• Cui Wei

Classifications

• H04L67/1004

  Server selection for load balancing · CPC title

• G06F9/5077 Primary

  Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455) · CPC title

• G06T1/20 Primary

  Processor architectures; Processor configuration, e.g. pipelining · CPC title

• G06T1/60

  Memory management · CPC title

• G06F9/5044

  considering hardware capabilities · CPC title