Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Resource scheduling method, resource scheduling system and equipment

A resource scheduling and resource technology, applied in the computer field, can solve the problems of reduced operating efficiency, resources cannot meet CRD resource requests, CRD cannot be scheduled, etc., and achieves the effect of high scheduling efficiency

Pending Publication Date: 2022-01-21
ZTE CORP
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, Kubernetes currently only supports Pod scheduling. To schedule CRDs requires a dedicated scheduler. Multiple schedulers will cause resource scheduling conflicts. At the same time, the following problems will also occur: resources cannot meet the resource requests of CRDs, resulting in CRDs being unable to be scheduled. ; Even if the CRD can be successfully scheduled, but the CRD is not scheduled according to the best resource allocation method, the operating efficiency will be reduced

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Resource scheduling method, resource scheduling system and equipment
  • Resource scheduling method, resource scheduling system and equipment
  • Resource scheduling method, resource scheduling system and equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0095] This embodiment is an example of the scheduler successfully scheduling CRDs and Pods. The embodiment shows the process of mixing CRDs and Pods on the Kubernetes scheduling platform. The deep learning job is defined as a CRD, and the workers who complete the deep learning job in parallel are assigned to the Pod. Bearer, which can realize the mixed scheduling of deep learning jobs and Pods and can run successfully.

[0096] Instance environment: a Kubernetes cluster equipped with Ubuntu 16.04 system, including two nodes with sufficient resources; the cluster has deployed a modified scheduler; deployed a custom deep learning job controller and splitter.

[0097] see Figure 7 , the specific operation steps are as follows:

[0098] Step S710: define a deep learning job file, and create the CRD object;

[0099] Step S720: define a single Pod file, and create the Pod object;

[0100] Step S730: After the deep learning job is successfully created, the CRD corresponding to t...

Embodiment 2

[0104] In this example, the scheduler successfully schedules two types of CRD objects. The example shows the process of mixing and scheduling different CRDs on the Kubernetes scheduling platform. The deep learning job is defined as a CRD, and the machine learning job is defined as a CRD, and the two types of CRD objects are executed. All Workers are carried by Pods, which can realize the mixed scheduling of deep learning jobs and machine learning jobs, and all of them can run successfully.

[0105] Instance environment: Kubernetes cluster equipped with Ubuntu 16.04 system, including two nodes with sufficient resources; the cluster has deployed a modified scheduler; deployed a custom deep learning job controller and splitter; deployed a custom machine Controller and splitter for learning jobs.

[0106] see Figure 8 , the specific operation steps are as follows:

[0107] Step S810: define the file of the deep learning job, and create the CRD object;

[0108] Step S820: defin...

Embodiment 3

[0115] In this embodiment, the scheduler schedules CRDs to run on the fewest nodes. The embodiment shows that when scheduling CRD objects on the Kubernetes scheduling platform, CRDs can be reasonably dismantled according to resource status, and deep learning jobs are defined as CRDs to complete deep learning. Workers that execute jobs in parallel are carried by Pods. When scheduling CRDs, the scheduler can automatically dismantle CRDs according to the current resource status, and schedule CRD Pods to run on extremely small nodes, reducing network overhead and ensuring the rationality of dismantling. .

[0116] Instance environment: Kubernetes cluster equipped with Ubuntu 16.04 system, including 3 nodes, the CPU and memory resources of the nodes are sufficient, node 1 has 8 idle GPUs, nodes 2 and 3 have 4 idle GPUs; the cluster has deployed the modified Scheduler; controller and splitter for deploying custom deep learning jobs.

[0117] see Figure 9 , the specific operation ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a resource scheduling method, a resource scheduling system, equipment and a computer readable storage medium, and the resource scheduling method comprises the steps: obtaining a scheduling object from a scheduling queue; when the scheduling object is a user-defined resource, disassembling the user-defined resource according to the current resource state, and obtaining a scheduling unit list, wherein the scheduling unit list comprises a first scheduling unit used for forming the user-defined resource; and sequentially scheduling the first scheduling units in the scheduling unit list. The method is suitable for a Kubernetes scheduling platform, if a scheduling object is CRD during scheduling, the CRD is disassembled according to a current resource state to obtain a scheduling unit list, the scheduling unit list comprises a Pod set, so that the Kubernetes scheduling platform can perform atomic scheduling on all Pods according to the scheduling unit list, all the Pods are scheduled in sequence according to a queue, insertion of other Pods is avoided, therefore, the CRD can be reasonably scheduled, the scheduling efficiency is high, and the Kubernetes scheduling platform can be compatible with various service scenes.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a resource scheduling method, a resource scheduling system, equipment, and a computer-readable storage medium. Background technique [0002] Kubernetes is currently the most mainstream container orchestration and scheduling platform. Kubernetes can support the management of user-defined resources (Custom Resource Definitions, CRD) through good scalability, which is convenient for users to manage custom resources as an overall object entity. However, Kubernetes currently only supports Pod scheduling. To schedule CRDs requires a dedicated scheduler. Multiple schedulers will cause resource scheduling conflicts. At the same time, the following problems will also occur: resources cannot meet the resource requests of CRDs, resulting in CRDs being unable to be scheduled. ; Even if the CRD can be successfully scheduled, but the CRD is not scheduled according to the optimal reso...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F9/50G06F9/48G06N3/08
CPCG06F9/5027G06F9/4881G06N3/08H04L67/1004H04L67/1095G06F9/505G06F9/48G06F9/50H04L67/1036G06F9/5038G06F9/5077G06F9/5061G06N20/00G06F2209/5017
Inventor 张乘铭唐波王科文韩炳涛王永成屠要峰高洪
Owner ZTE CORP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com