TWI894013B - Resource allocation method and system - Google Patents

Abstract

The present disclosure proposes a resource allocation method and system. The resource allocation method includes: determining a target resource requirement from candidate resource requirements according to a type of a designated job included in an input command, obtaining total available hardware resource according to occupied hardware resource of existing jobs and current available hardware resource when the current available hardware resource does not satisfy the target resource requirement and priorities of the existing jobs are lower than a priority of the designated job, allocating target resource satisfying the target resource requirement from the total available hardware resource to the designated job, and allocating first resource to a first existing job and second resource to a second existing job, respectively, when remaining resource of the total available hardware resource does not satisfy resource requirements of the existing jobs, wherein the first resource is lower than the second resource, and a priority of the first existing job is lower than a priority of the second existing job.

Description

Resource allocation method and system

The present invention relates to a resource allocation method and system.

In traditional software development, developers typically use personal computers as their development platform. However, as project sizes continue to grow, the performance of personal computers is increasingly insufficient. Therefore, the industry often uses more powerful workstations or servers as development platforms.

However, when the development environment migrates from a personal computer to a shared server, uneven hardware resource allocation often occurs, impacting development efficiency and work progress.

In view of the above, the present invention provides a resource allocation method and system to solve the above problems.

According to an embodiment of the present invention, a resource allocation method is executed by a processing device, comprising: determining a target resource requirement from a plurality of candidate resource requirements according to the type of a designated task included in an input command; when the currently available hardware resources do not meet the target resource requirement and the priorities of the plurality of existing tasks are lower than the priority of the designated task, reclaiming the first occupied hardware resources of the existing tasks to obtain a first available total hardware resource based on the first occupied hardware resources and the currently available hardware resources; and allocating the first available total hardware resources that meet the target resource requirement to the first available total hardware resource. allocating the requested target resource to the designated task; and when first remaining hardware resources excluding the target resource of the first available total hardware resources do not meet the resource requirements of the existing tasks, allocating a first resource from the first remaining hardware resources to a first existing task among the existing tasks, and allocating a second resource excluding the first resource from the first remaining hardware resources to at least one second existing task among the existing tasks, wherein a priority of the first existing task is lower than a priority of the at least one second existing task, and wherein the second resource is higher than the first resource.

According to an embodiment of the present invention, a resource allocation system includes: a memory device and a processing device. The memory device is used to store multiple candidate resource requirements. The processing device is connected to the memory device, and the processing device is used to execute: judging the target resource requirement from multiple candidate resource requirements according to the type of designated work included in the input instruction; when the currently available hardware resources do not meet the target resource requirement and the multiple priorities of multiple existing jobs are lower than the priority of the designated job, reclaiming the first occupied hardware resources of these existing jobs to obtain the first available total hardware resources based on the first occupied hardware resources and the currently available hardware resources; and allocating the target resource requirements that meet the target resource requirement from the first available total hardware resources. allocating target resources to designated tasks; and when first remaining hardware resources excluding the target resources of the first available total hardware resources do not meet the resource requirements of the existing tasks, allocating first resources from the first remaining hardware resources to a first existing task among the existing tasks, and allocating second resources excluding the first resources from the first remaining hardware resources to at least one second existing task among the existing tasks, wherein a priority of the first existing task is lower than a priority of the at least one second existing task, and wherein the second resource is higher than the first resource.

In summary, the resource allocation method and system according to one or more of the above embodiments can avoid the problem of uneven hardware resource allocation and ensure that more important tasks are preferentially allocated sufficient hardware resources. Therefore, whether migrating tasks from a personal computer to a shared server or migrating tasks from a server to a personal computer, hardware resources can be effectively allocated.

The above description of the present disclosure and the following description of the embodiments are intended to demonstrate and explain the spirit and principles of the present invention, and to provide further explanation of the scope of the patent application of the present invention.

The following detailed description of the features and advantages of the present invention is sufficient to enable anyone skilled in the relevant art to understand the technical content of the present invention and implement it accordingly. Based on the disclosure, patent application scope, and drawings in this specification, anyone skilled in the relevant art can easily understand the relevant objectives and advantages of the present invention. The following examples are intended to further illustrate the concepts of the present invention in detail and are not intended to limit the scope of the present invention in any way.

Please refer to FIG1 , which is a block diagram of a resource allocation system according to an embodiment of the present invention. As shown in FIG1 , the resource allocation system 1 includes a memory device 11 and a processing device 12. The memory device 11 is communicatively or electrically connected to the processing device 12.

The memory device 11 is used to store multiple candidate resource requirements. The candidate resource requirements may correspond to different work types, and each work type may correspond to different hardware resource configuration information. Work types may include multi-threaded compilation (e.g., C language programs), compilation scenarios requiring simultaneous code downloading during compilation (e.g., OpenBMC projects), data processing, and artificial intelligence training. Hardware resource configuration information corresponding to multi-threaded compilation may include: a 64-core CPU, 8GB of DRAM, a 50GB solid-state drive, and a 0-core graphics processor. The hardware resource configuration information for a compilation scenario requiring simultaneous code download may include: a 32-core CPU, 32GB of RAM, a 100GB SSD, and a 0-core GPU. The hardware resource configuration information for data processing and AI training may include: an 8-core CPU, 12GB of RAM, a 50GB SSD, and a 64-core GPU. The above workload types and hardware resource configuration information are examples only and are not intended to limit the present invention. The memory device 11 may include one or more memories, which may be non-volatile memories (NVM), such as read-only memory (ROM), flash memory, and/or non-volatile random access memory (NVRAM).

The processing device 12 is used to allocate hardware resources based on the type of designated task and the corresponding resource requirements in the candidate resource requirements. These hardware resources can be those of a computing device such as a personal computer or a server. The following description uses a server as an example, but the present invention is not limited thereto. The processing device 12 can include one or more processors, such as a central processing unit (CPU), a graphics processor (GPU), a microcontroller (MCU), a programmable logic controller (PLC), or other processors with signal processing capabilities.

Please refer to Figures 1 and 2 together, wherein Figure 2 is a flow chart of a resource allocation method according to an embodiment of the present invention. The resource allocation method can be executed by the processing device 12 shown in Figure 1. As shown in Figure 2, the resource allocation method includes: Step S101: determining a target resource requirement from a plurality of candidate resource requirements based on the type of designated work included in the input command; Step S103: when the currently available hardware resources do not meet the target resource requirement and the plurality of priorities of the plurality of existing jobs are lower than the priority of the designated work, reclaiming the first occupied hardware resources of the existing jobs to obtain a first available total hardware resource based on the first occupied hardware resources and the currently available hardware resources; Step S105: Allocating target resources that meet target resource requirements from the first available total hardware resources to the designated task; and step S107: when first remaining hardware resources excluding the target resources from the first available total hardware resources do not meet resource requirements of the existing tasks, allocating first resources from the first remaining hardware resources to a first existing task among the existing tasks, and allocating second resources excluding the first resources from the first remaining hardware resources to at least one second existing task among the existing tasks.

In step S101, processing device 12 receives an input command from a user, which may include a designated task type. Based on the designated task type, processing device 12 selects one of the candidate resource requirements stored in memory device 11 as a target resource requirement. For example, if the designated task type is multi-threaded compilation, the target resource requirements are a 64-core CPU, 8GB of DRAM, a 50GB SSD, and a 0-core graphics processor.

In step S103, when the processing device 12 determines that the server's currently available hardware resources do not meet the target resource requirements and that the priorities of all existing tasks currently being executed by the server are lower than the priority of the designated task, the processing device 12 reclaims the first occupied hardware resources of these existing tasks to obtain a first total available hardware resource based on the first occupied hardware resources and the currently available hardware resources. The first occupied hardware resource may be the sum of the hardware resources used to execute all existing tasks, and the currently available hardware resources may be idle resources. Taking the central processing unit (CPU) as an example, if the currently available hardware resources indicate that the number of available cores of the CPU is zero, the first total available hardware resources are the first occupied hardware resources. If the currently available hardware resources indicate that the number of available cores of the CPU is greater than zero, the first total available hardware resources are the sum of the first occupied hardware resources and the currently available hardware resources.

In step S105, the processing device 12 allocates target resources from the first available total hardware resources that meet the target resource requirements of the designated task, based on the designated task's higher priority than existing tasks. Furthermore, the target resources may be allocated to a container for the designated task, which may then be activated to execute the designated task.

In step S107, after allocating the target resource to the designated task, when the processing device 12 determines that the first remaining hardware resources of the server's first total available hardware resources excluding the target resource do not meet the resource requirements of the existing tasks, and the priority of the first existing task among the existing tasks is lower than the priority of at least one second existing task among the existing tasks, the processing device 12 allocates the first resource from the first remaining hardware resources to the first existing task, and allocates the second resource excluding the first resource from the first remaining hardware resources to the second existing task, where the second resource is higher than the first resource.

Specifically, after allocating target resources from the first available total hardware resources to a designated task, the remaining hardware resources on the server are first remaining hardware resources. If the first remaining hardware resources cannot meet the resource requirements of all existing tasks, the processing device 12 allocates first resources from the first remaining hardware resources to the first existing task and allocates remaining second resources from the first remaining hardware resources, excluding the first resources, to a second existing task. The first resources may be minimum allocation resources. In one embodiment, if there are multiple second existing tasks, the processing device 12 may evenly distribute the second resources among the second existing tasks.

Thus, the resource allocation method and system according to one or more of the above embodiments can avoid the problem of uneven hardware resource allocation and ensure that more important tasks are preferentially allocated sufficient hardware resources. Therefore, whether migrating tasks from a personal computer to a shared server or migrating tasks from a server to a personal computer, hardware resources can be effectively allocated.

Please refer to FIG. 1 and FIG. 3 , wherein FIG. 3 is a flowchart illustrating a resource allocation method according to an embodiment of the present invention for allocating resources to a task whose priority is not higher than that of an existing task. The steps shown in FIG. 3 may be performed by the processing device 12 shown in FIG. 1 . As shown in FIG3 , the resource allocation method further includes: step S201: when currently available hardware resources do not meet target resource requirements and the priorities of the existing tasks are not lower than the priority of the designated task, reclaiming the second occupied hardware resources of the existing tasks to obtain a second total available hardware resource based on the second occupied hardware resources and the currently available hardware resources; step S203: allocating at least one third resource from the second total available hardware resources to at least one third existing task having the highest priority among the priorities among the existing tasks; and step S205: allocating a portion of the second remaining hardware resources excluding the at least one third resource from the second total available hardware resources to the designated task and at least one fourth resource to at least one fourth existing task.

After the processing device 12 receives the input command from the user, in step S201, if the processing device 12 determines that the server's currently available hardware resources do not meet the target resource requirements and the priority of existing jobs being executed by the server is not lower than the priority of the designated job, the processing device 12 reclaims the second occupied hardware resources of these existing jobs to obtain a second total available hardware resource based on the second occupied hardware resources and the currently available hardware resources. The second occupied hardware resources may be the sum of the hardware resources used to execute all existing jobs, and the currently available hardware resources may be idle resources. Taking the central processing unit (CPU) as an example, if the currently available hardware resources indicate that the number of available cores of the CPU is zero, the second total available hardware resources are the second occupied hardware resources. If the currently available hardware resources indicate that the number of available cores of the CPU is greater than zero, the second total available hardware resources are the sum of the second occupied hardware resources and the currently available hardware resources.

In step S203, the processing device 12 allocates a third resource from the second available total hardware resources that satisfies the third existing task to the third existing task based on the third existing task having the highest priority among the existing tasks. Furthermore, the third resource may be allocated to a container for the third existing task, and the container for the third existing task may be activated to execute the third existing task.

In step S205, after allocating the third resource to the third existing task, the processing device 12 allocates a portion of the hardware resources of the second available total hardware resources excluding the third resource to the designated task, and allocates the fourth resource to the fourth existing task among the existing tasks.

Specifically, after allocating the third resource from the second available total hardware resources to the third existing job with the highest priority, the remaining hardware resources on the server are the second remaining hardware resources. When the priority of the designated job is lower than the priority of the fourth existing job, the portion of resources allocated by processing device 12 to the designated job is lower than the fourth resource allocated to the fourth existing job. When the priority of the designated job is equal to the priority of the fourth existing job, the portion of resources allocated by processing device 12 to the designated job is equal to the fourth resource allocated to the fourth existing job.

In one embodiment, when both the priority of the designated task and the priority of the third existing task are the highest, the processing device 12 may allocate the target resource requirements of the designated task from the second available total hardware resources to the designated task and allocate the hardware resources required to meet the requirements of the third existing task to the third existing task. If the remaining hardware resources are insufficient to meet the requirements of the other existing tasks, the processing device 12 may allocate the minimum allocation resources to one or more of the existing tasks with the lowest priority and evenly distribute the remaining hardware resources among the remaining existing tasks.

In addition, when the priority of the designated task and the priority of the third existing task are both the highest priority and the second total available hardware resources do not meet the target resource requirements and the resource requirements of the third existing task, the processing device 12 may evenly distribute the second total available hardware resources to the designated task and the third existing task.

It should be noted that the number of the third and fourth existing tasks can be one or more, and the present invention does not limit this. Furthermore, depending on the hardware resources occupied by the existing tasks, the first and second occupied hardware resources can be the same or different, and the present invention does not limit this.

Furthermore, the resource allocation method may further include determining the priority of a designated task based on a user instruction. Specifically, the user instruction may indicate the priority of the designated task. The processing device 12 may receive the user instruction and execute step S103 of FIG. 2 and step S201 of FIG. 3 based on the priority indicated in the user instruction. Furthermore, the priority of an existing task may also be determined based on the user instruction.

In addition, in the resource allocation method and system of one or more of the above embodiments, when a new designated task is added or an existing task is completed, the processing device 12 may generate an interrupt to reallocate resources to execute step S103 of the figure and/or step S201 of Figure 3 again.

In summary, the resource allocation method and system according to one or more of the above embodiments can avoid the problem of uneven hardware resource allocation and ensure that more important tasks are preferentially allocated sufficient hardware resources. Therefore, whether migrating tasks from a personal computer to a shared server or migrating tasks from a server to a personal computer, hardware resources can be effectively allocated.

Although the present invention is disclosed above with reference to the aforementioned embodiments, they are not intended to limit the present invention. Any modifications and enhancements that do not depart from the spirit and scope of the present invention are within the scope of patent protection of the present invention. Please refer to the attached patent application for the scope of protection defined by the present invention.

1: Resource Allocation System 11: Memory Device 12: Processing Device S101, S103, S105, S107, S201, S203, S205: Steps

Figure 1 is a block diagram of a resource allocation system according to an embodiment of the present invention. Figure 2 is a flow chart of a resource allocation method according to an embodiment of the present invention. Figure 3 is a flow chart of a resource allocation method according to an embodiment of the present invention, illustrating a method for allocating resources in which the priority of a designated task is no higher than that of an existing task.

S101, S103, S105, S107: Steps

Claims (8)

A resource allocation method is executed by a processing device, comprising: determining a target resource requirement from a plurality of candidate resource requirements according to a type of a designated task included in an input command; when a currently available hardware resource does not meet the target resource requirement and a plurality of priorities of a plurality of existing tasks are lower than the priority of the designated task, reclaiming a first occupied hardware resource of the existing tasks to obtain a first available total hardware resource based on the first occupied hardware resource and the currently available hardware resource; allocating the first available total hardware resource that meets the target resource requirement to the first available total hardware resource; allocating a target resource of the target resource requirement to the designated task; and when a first remaining hardware resource of the first available total hardware resources excluding the target resource does not satisfy the resource requirements of the existing tasks, allocating a first resource from the first remaining hardware resource to a first existing task among the existing tasks, and allocating a second resource of the first remaining hardware resource excluding the first resource to at least one second existing task among the existing tasks, wherein the priority of the first existing task is lower than the priority of the at least one second existing task and wherein the second resource is higher than the first resource, wherein the resource allocation method further comprises: when the currently available hardware resources do not meet the target resource requirement and the priorities of the existing tasks are not lower than the priority of the designated task, reclaiming a second occupied hardware resource of the existing tasks to obtain a second available total hardware resource based on the second occupied hardware resource and the currently available hardware resource; allocating at least a third resource from the second available total hardware resource to the task with the highest priority among the existing tasks; at least one third existing job with a priority level; and allocating a portion of resources to the designated job and at least one fourth resource to at least one fourth existing job from a second remaining hardware resource excluding the at least one third resource in the second available total hardware resources, wherein when the priority level of the designated job is lower than the priority level of the at least one fourth existing job, the portion of resources is lower than the at least one fourth resource, and when the priority level of the designated job is equal to the priority level of the at least one fourth existing job, the portion of resources is equal to the at least one fourth resource. The resource allocation method as described in claim 1, wherein the at least one second existing task is a plurality of second existing tasks, and allocating the second resources other than the first resources in the first remaining hardware resources to the at least one second existing task among the existing tasks includes: evenly allocating the second resources to the second existing tasks. The resource allocation method as described in claim 1 further includes: determining the priority level of the designated task based on a user instruction. The resource allocation method of claim 1, wherein the type of the designated task includes at least one of multi-threaded compilation, simultaneous compilation and code downloading, data processing, and artificial intelligence training. A resource allocation system includes: a memory device for storing a plurality of candidate resource requirements; and a processing device connected to the memory device, the processing device being configured to: determine a target resource requirement from the candidate resource requirements based on a type of a designated task included in an input instruction; and when a currently available hardware resource does not meet the target resource requirement and a plurality of priorities of a plurality of existing tasks are lower than the priority of the designated task, reclaim a first occupied hardware resource of the existing tasks to obtain a target resource requirement based on the first occupied hardware resource and the currently available hardware resource. a first available total hardware resource; allocating a target resource that meets the target resource requirement from the first available total hardware resource to the designated task; and when a first remaining hardware resource of the first available total hardware resource excluding the target resource does not meet the resource requirements of the existing tasks, allocating a first resource from the first remaining hardware resource to a first existing task among the existing tasks, and allocating a second resource from the first remaining hardware resource excluding the first resource to at least one second existing task among the existing tasks, wherein the priority of the first existing task is the priority of the first existing task. The priority of the at least one second existing job is lower than the priority of the at least one second existing job, and the second resource is higher than the first resource, wherein the processing device is further configured to: when the currently available hardware resources do not meet the target resource requirement and the priorities of the existing jobs are not lower than the priority of the designated job, reclaim a second occupied hardware resource of the existing jobs to obtain a second available total hardware resource based on the second occupied hardware resource and the currently available hardware resource; and allocate at least one third resource from the second available total hardware resource to the existing jobs having the second occupied hardware resource. at least one third existing job of the highest priority among the priorities; and allocating a portion of resources to the designated job and at least one fourth resource to at least one fourth existing job from a second remaining hardware resource excluding the at least one third resource in the second available total hardware resources, wherein when the priority of the designated job is lower than the priority of the at least one fourth existing job, the portion of resources is lower than the at least one fourth resource, and when the priority of the designated job is equal to the priority of the at least one fourth existing job, the portion of resources is equal to the at least one fourth resource. The resource allocation system as described in claim 5, wherein the at least one second existing task is a plurality of second existing tasks, and the processing device evenly distributes the second resource to the second existing tasks. The resource allocation system as described in claim 5, wherein the processing device is further configured to determine the priority of the designated task based on a user instruction. The resource allocation system of claim 5, wherein the type of the designated task includes at least one of multi-threaded compilation, concurrent compilation and code downloading, data processing, and artificial intelligence training.