CN116302347A - Heterogeneous container cluster self-adaptive scheduling method and device - Google Patents

Heterogeneous container cluster self-adaptive scheduling method and device Download PDF

Info

Publication number
CN116302347A
CN116302347A CN202310263708.5A CN202310263708A CN116302347A CN 116302347 A CN116302347 A CN 116302347A CN 202310263708 A CN202310263708 A CN 202310263708A CN 116302347 A CN116302347 A CN 116302347A
Authority
CN
China
Prior art keywords
node
pod
weight value
heterogeneous
adaptive scheduling
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310263708.5A
Other languages
Chinese (zh)
Inventor
朱廷祥
王群
张强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianyi Cloud Technology Co Ltd
Original Assignee
Tianyi Cloud Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianyi Cloud Technology Co Ltd filed Critical Tianyi Cloud Technology Co Ltd
Priority to CN202310263708.5A priority Critical patent/CN116302347A/en
Publication of CN116302347A publication Critical patent/CN116302347A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • G06F9/45558—Hypervisor-specific management and integration aspects
    • G06F2009/45595—Network integration; Enabling network access in virtual machine instances
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a heterogeneous container cluster self-adaptive scheduling method, a heterogeneous container cluster self-adaptive scheduling device and a storage medium, and relates to the field of high performance of computer technology. The heterogeneous container cluster self-adaptive scheduling method comprises the following steps: inquiring the current resource utilization rate of each node when the container scheduling platform senses service change; when pod needs to expand, according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, calculating the node with the highest expansion weight value, and then expanding the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy; when pod needs to shrink, according to the current resource utilization rate and a weighted average algorithm, calculating the node with the highest shrink weight value, and then combining with the pod affinity strategy, prioritizing the node with the highest shrink weight value and conforming to the affinity strategy. The invention integrates the stability of the heterogeneous container cluster and solves the problem of low self-adaptive scheduling performance in the heterogeneous cluster.

Description

Heterogeneous container cluster self-adaptive scheduling method and device
Technical Field
The present invention relates to the field of high performance computer technologies, and in particular, to a heterogeneous container cluster adaptive scheduling method, apparatus and storage medium.
Background
Currently, cloud computing technology has gradually transitioned from a virtual machine technology architecture represented by VMM and Openstack to a cloud native technology architecture represented by container and K8S, which has a relation with the increasing complexity of application programs. Traditional application programs are all single-body architectures, and as the programs become more and more complex, the single-body architectures are split into distributed architectures of a plurality of services, and gradually evolved into later micro-service architectures. The container is more suitable for a system of a distributed architecture because of the small occupation of resources, and under the background of the requirement, the container and the cloud native technology represented by K8S become the main stream of cloud computing. Meanwhile, under the dual driving of international situation and policy, the creation cloud construction is gradually accelerated, and autonomous IT architecture and standard are gradually established starting from autonomous innovation of key software and hardware components such as chips, servers, cloud platforms, operating systems, databases, middleware and the like. The created container cloud is taken as an important branch of the created cloud, and has wide application scenes in industries such as government and finance by combining the advantages of the container cloud architecture.
In a belief-creation cloud scene, a plurality of domestic chips are generally adopted to form a cloud resource pool at the bottom layer, and in a virtualization layer, the application of the cloud resource pool to the chip adaptation of different architectures can bring great challenges. The container technology is constructed on an operating system, belongs to process-level virtualization, and can shield architecture differences and realize heterogeneous unified container cloud resource pools.
However, in the process of implementing the technical scheme of the invention in the embodiment of the application, the inventor of the application finds that at least the following technical problems exist in the above technology:
in the prior art, in the process of unified scheduling in a container resource pool, the great difference of calculation power and resource capacity of different domestic chips is not considered, so that the self-adaptive scheduling performance of the information-created container cluster in a heterogeneous scene under the construction of domestic multiple chips is low.
Disclosure of Invention
The embodiment of the application solves the problem of low self-adaptive scheduling performance of the information-created container cluster in a heterogeneous scene under the domestic multiple chip architectures by providing the self-adaptive scheduling method, the self-adaptive scheduling device and the storage medium for the heterogeneous container cluster, improves the self-adaptive scheduling capability, and enables the self-adaptive scheduling to be more accurate.
The embodiment of the application provides a heterogeneous container cluster self-adaptive scheduling method, which comprises the following steps:
s1, inquiring the current resource utilization rate of each node when a container scheduling platform senses service change;
s2, when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the pod to the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and S3, when the pod needs to shrink, calculating a node with the highest shrink weight value according to the current resource utilization rate and a weighted average algorithm, and then, combining the pod affinity strategy, and preferentially shrinking the node with the highest shrink weight value and conforming to the affinity strategy.
Furthermore, the container dispatching platform is a container cluster under a domestic multiple chip architecture.
Furthermore, the calculation coefficients of the chips are different, 1 is taken as a reference by intel 5218 single-core performance, and the equivalent percentages of other chips are calculation coefficients.
Further, the service change in S1 includes increasing the workload and decreasing the workload, i.e. the number of pod.
Further, the current resource usage in S2 and S3 includes resource usage of CPU, memory, disk, and network.
Further, the node basic attribute in S2 includes a CPU dominant frequency and a core number.
Further, the weighted average algorithm in S2 and S3 includes:
expansion weight value= [ power coefficient x CPU core number x (1-CPU usage) ×cpu weight+idle memory x memory weight ] ×if (pod memory requirement > idle memory, 0, 1);
the reduction weight value= [ CPU usage x CPU weight+memory usage x memory weight ].
Further, when the capacity is expanded in the capacity expansion weight value, the available idle memory is compared with the pod memory requirement, if the idle memory is insufficient to support the pod memory requirement, the capacity expansion weight of the whole node is set to 0, otherwise, if the idle memory is sufficient, the capacity expansion weight of the whole node is set to 1.
The embodiment of the application provides a heterogeneous container cluster self-adaptive scheduling device, which comprises a query module, a capacity expansion module and a capacity contraction module:
and a query module: the method comprises the steps that when a container scheduling platform senses service change, the current resource utilization rate of each node is inquired;
and the capacity expansion module is used for: when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and a capacity shrinking module: when the pod needs to shrink, according to the current resource utilization rate and the weighted average algorithm, the node with the highest shrink weight value is calculated, and then the node with the highest shrink weight value and conforming to the affinity strategy is preferentially shrunk by combining with the pod affinity strategy.
The embodiment of the application provides a computer readable storage medium for storing a program which, when executed by a processor, implements a heterogeneous container cluster adaptive scheduling method.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
1. because a weighted average algorithm is adopted, the expansion and contraction capacity weight value of the container cluster scheduling is calculated according to the factors such as the CPU chip computing power, the weight value of the container scheduling is quantized, the problem that the self-adaptive scheduling performance of the information-created container cluster in heterogeneous scenes under the domestic multiple chip architectures in the prior art is low is effectively solved, the self-adaptive scheduling capability is improved, and the self-adaptive scheduling is more accurate.
2. Because the technology taking 1 as a reference and the equivalent percentage of other chips as the power coefficient is adopted by the single-core performance of intel 5218, different chips have uniform power standards, the problem of low scheduling performance caused by the power difference under different chip architectures in a container cluster under a cross-architecture in the prior art is effectively solved, and the problem of cross-platform deployment scheduling of the container cluster is further realized.
3. Because the resources and calculation forces of different nodes under the heterogeneous container cluster are comprehensively considered and used as factors for calculating the scheduling weight value, the resource balancing can be carried out on the different nodes, the problem of low resource balancing capability of different node nodes of the heterogeneous container cluster in the prior art is effectively solved, and the stability of integrating the heterogeneous container cluster is further realized.
Drawings
Fig. 1 is a flowchart of a heterogeneous container cluster adaptive scheduling method according to an embodiment of the present application;
FIG. 2 is a graph of a conventional chip power coefficient according to an embodiment of the present disclosure;
FIG. 3 is a schematic diagram of a heterogeneous clustered pod adaptive schedule according to an embodiment of the present disclosure;
fig. 4 is a block diagram of a heterogeneous container cluster adaptive scheduling device according to an embodiment of the present application.
Detailed Description
According to the self-adaptive scheduling method, device and storage medium for the heterogeneous container clusters, the problem that the self-adaptive scheduling performance of the information-created container clusters in heterogeneous scenes is low in the prior art under the condition of localization of various chip architectures is solved, the self-adaptive scheduling of pod is achieved by means of a weighted average algorithm in the scheduling process of the heterogeneous container clusters, self-adaptive scheduling capability is improved, and the self-adaptive scheduling is more accurate.
The technical scheme in the embodiment of the application aims to solve the problem of low self-adaptive scheduling performance of the information creation container cluster in heterogeneous scenes under the domestic multiple chip architectures, and the overall thought is as follows:
inquiring the current resource utilization rate of each node when the container scheduling platform senses that the service needs to increase or decrease the workload; when pod needs to expand, according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, calculating the node with the highest expansion weight value, and then expanding the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy; when the pod needs to shrink, according to the current resource utilization rate and a weighted average algorithm, the node with the highest shrink weight value is calculated, and then the node with the highest shrink weight value and conforming to the affinity policy is preferentially selected by combining with the pod affinity policy, so that the resource balancing capability of different node nodes of the heterogeneous container cluster is effectively improved, and the stability of integrating the heterogeneous container cluster is realized.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Example 1
As shown in fig. 1, a flowchart of a heterogeneous container cluster adaptive scheduling method provided in an embodiment of the present application is applied to a heterogeneous container cluster adaptive scheduling device, where the method includes the following steps:
s1, inquiring the current resource utilization rate of each node when a container scheduling platform senses service change;
s2, when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the pod to the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and S3, when the pod needs to shrink, calculating a node with the highest shrink weight value according to the current resource utilization rate and a weighted average algorithm, and then, combining the pod affinity strategy, and preferentially shrinking the node with the highest shrink weight value and conforming to the affinity strategy.
Further, the container dispatching platform is a container cluster under a plurality of domestic chip architectures.
In this embodiment, the container scheduling platform is a credit container cluster under a domestic multiple chip architecture.
Furthermore, the calculation coefficients of the various chips are different, 1 is taken as a reference by intel 5218 single-core performance, and the equivalent percentages of other chips are calculation coefficients.
In this embodiment, as shown in the common chip power coefficient diagram shown in fig. 2, other chips can be converted according to the above standard to obtain the related power coefficient parameters, although not shown by way of example.
Further, the traffic change in S1 includes increasing the workload and decreasing the workload, i.e., the number of pod.
Further, the current resource usage in S2 and S3 includes resource usage of CPU, memory, disk, network.
Further, the node basic attributes in S2 include the CPU dominant frequency and the core number.
In this embodiment, the expansion and contraction capacity weight value of the container cluster scheduling is calculated according to factors such as the calculation power of the CPU chip and the memory, so as to realize accurate self-adaptive scheduling capability.
Further, the weighted average algorithm in S2 and S3 includes:
expansion weight value= [ power coefficient x CPU core number x (1-CPU usage) ×cpu weight+idle memory x memory weight ] ×if (pod memory requirement > idle memory, 0, 1);
the reduction weight value= [ CPU usage x CPU weight+memory usage x memory weight ].
And when the capacity is expanded in the capacity expansion weight value, comparing the available idle memory with the pod memory requirement, setting the capacity expansion weight of the whole node to 0 if the idle memory is insufficient to support the pod memory requirement, otherwise, setting the capacity expansion weight of the whole node to 1 if the idle memory is sufficient.
In this embodiment, as shown in the pod adaptive scheduling diagram under the heterogeneous cluster in fig. 3, only the utilization factor of the CPU and the memory as a whole is considered when the pod is scaled or deleted, and the higher the utilization, the larger the weight, and the priority is scaled.
As shown in fig. 4, in order to provide a structure diagram of a heterogeneous container cluster adaptive scheduling device according to an embodiment of the present application, the heterogeneous container cluster adaptive scheduling device provided by the embodiment of the present application includes a query module, an expansion module, and a contraction module:
and a query module: the method comprises the steps that when a container scheduling platform senses service change, the current resource utilization rate of each node is inquired;
and the capacity expansion module is used for: when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and a capacity shrinking module: when the pod needs to shrink, according to the current resource utilization rate and the weighted average algorithm, the node with the highest shrink weight value is calculated, and then the node with the highest shrink weight value and conforming to the affinity strategy is preferentially shrunk by combining with the pod affinity strategy.
In the embodiment, firstly, when a query module senses service change in a container scheduling platform, the current resource utilization rate of each node is queried; then when the pod needs to expand in the expansion module, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the pod to the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy; and when the pod needs to be scaled in the scaling module, calculating a node with the highest scaling weight value according to the current resource utilization rate and a weighted average algorithm, and then, combining the pod affinity strategy, and prioritizing the node with the highest scaling weight value and conforming to the affinity strategy.
The embodiment of the application also provides a computer readable storage medium for storing a program, and the program is executed by a processor to realize the heterogeneous container cluster adaptive scheduling method.
The technical scheme in the embodiment of the application at least has the following technical effects or advantages: the weighted average algorithm is adopted to the container scheduling in the heterogeneous cluster to energy the weight value of the container scheduling, so that the self-adaptive scheduling of the container is realized, the problems of low scheduling policy and mechanism efficiency of the containers with different chip architectures are solved, the development potential of the credit-invasive container can be accelerated, and the cross-platform deployment scheduling of the container cluster is realized.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims (10)

1. The adaptive scheduling method for the heterogeneous container cluster is characterized by comprising the following steps of:
s1, inquiring the current resource utilization rate of each node when a container scheduling platform senses service change;
s2, when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the pod to the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and S3, when the pod needs to shrink, calculating a node with the highest shrink weight value according to the current resource utilization rate and a weighted average algorithm, and then, combining the pod affinity strategy, and preferentially shrinking the node with the highest shrink weight value and conforming to the affinity strategy.
2. The heterogeneous container cluster adaptive scheduling method of claim 1, wherein: the container dispatching platform is a container cluster under a plurality of domestic chip architectures.
3. The heterogeneous container cluster adaptive scheduling method of claim 2, wherein: the calculation coefficients of the chips are different, 1 is taken as a reference by the single-core performance of intel 5218, and the equal percentages of other chips are calculation coefficients.
4. The heterogeneous container cluster adaptive scheduling method of claim 1, wherein: the service change in S1 includes increasing the workload and decreasing the workload, i.e. the number of pod.
5. The heterogeneous container cluster adaptive scheduling method of claim 1, wherein: the current resource utilization rate in S2 and S3 comprises the resource utilization conditions of CPU, memory, disk and network.
6. The heterogeneous container cluster adaptive scheduling method of claim 1, wherein: the node basic attributes in the S2 comprise a CPU main frequency and a core number.
7. The adaptive scheduling method of heterogeneous container clusters according to claim 1, wherein the weighted average algorithm in S2 and S3 comprises:
expansion weight value= [ power coefficient x CPU core number x (1-CPU usage) ×cpu weight+idle memory x memory weight ] ×if (pod memory requirement > idle memory, 0, 1);
the reduction weight value= [ CPU usage x CPU weight+memory usage x memory weight ].
8. The heterogeneous container cluster adaptive scheduling method of claim 7, wherein: and when the capacity is expanded in the capacity expansion weight value, comparing the available idle memory with the pod memory requirement, setting the capacity expansion weight of the whole node to 0 if the idle memory is insufficient to support the pod memory requirement, otherwise, setting the capacity expansion weight of the whole node to 1 if the idle memory is sufficient.
9. The self-adaptive scheduling device for the heterogeneous container clusters is characterized by comprising a query module, a capacity expansion module and a capacity contraction module:
and a query module: the method comprises the steps that when a container scheduling platform senses service change, the current resource utilization rate of each node is inquired;
and the capacity expansion module is used for: when the pod needs to expand, calculating a node with the highest expansion weight value according to the current resource utilization rate, the node basic attribute and the weighted average algorithm, and expanding the node with the highest weight value and conforming to the affinity strategy by combining with the pod affinity strategy;
and a capacity shrinking module: when the pod needs to shrink, according to the current resource utilization rate and the weighted average algorithm, the node with the highest shrink weight value is calculated, and then the node with the highest shrink weight value and conforming to the affinity strategy is preferentially shrunk by combining with the pod affinity strategy.
10. A computer readable storage medium storing a program, wherein the program when executed by a processor implements the heterogeneous container cluster adaptive scheduling method of any one of claims 1 to 8.
CN202310263708.5A 2023-03-10 2023-03-10 Heterogeneous container cluster self-adaptive scheduling method and device Pending CN116302347A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310263708.5A CN116302347A (en) 2023-03-10 2023-03-10 Heterogeneous container cluster self-adaptive scheduling method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310263708.5A CN116302347A (en) 2023-03-10 2023-03-10 Heterogeneous container cluster self-adaptive scheduling method and device

Publications (1)

Publication Number Publication Date
CN116302347A true CN116302347A (en) 2023-06-23

Family

ID=86801041

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310263708.5A Pending CN116302347A (en) 2023-03-10 2023-03-10 Heterogeneous container cluster self-adaptive scheduling method and device

Country Status (1)

Country Link
CN (1) CN116302347A (en)

Similar Documents

Publication Publication Date Title
CN111221624B (en) Container management method for regulation and control cloud platform based on Docker container technology
CN111339027B (en) Automatic design method of reconfigurable artificial intelligent core and heterogeneous multi-core chip
CN110058924A (en) A kind of container dispatching method of multiple-objection optimization
CN106412124B (en) A kind of and sequence cloud service platform task distribution system and method for allocating tasks
CN116414559A (en) Method for modeling and distributing unified computing power identification, storage medium and electronic equipment
CN109040156B (en) Soft load resource processing method and device based on container cloud
CN110134646B (en) Knowledge platform service data storage and integration method and system
CN116185629A (en) Task execution method and device, storage medium and electronic equipment
CN113645076B (en) Virtual network resource allocation method based on hypergraph matching algorithm
CN114339891A (en) Edge unloading resource allocation method and system based on Q learning
CN113360576A (en) Power grid mass data real-time processing method and device based on Flink Streaming
CN112799829A (en) Knowledge-driven network resource arrangement method
CN112965788A (en) Task execution method, system and equipment in hybrid virtualization mode
CN116302347A (en) Heterogeneous container cluster self-adaptive scheduling method and device
CN107948330A (en) Load balancing based on dynamic priority under a kind of cloud environment
CN115051998A (en) Adaptive edge computing offloading method, apparatus and computer-readable storage medium
CN110908800B (en) Scheduling optimization method based on scheduling entropy in edge cloud collaborative environment
CN115617526A (en) Cloud data center energy-saving method based on cloud data center construction and virtual machine integration
CN112148496B (en) Energy efficiency management method and device for computing storage resources of super-fusion virtual machine and electronic equipment
Swarnakar et al. A novel improved hybrid model for load balancing in cloud environment
US20090249020A1 (en) Techniques for optimizing configuration partitioning
CN117519996B (en) Data processing method, device, equipment and storage medium
CN113127205B (en) Workflow scheduling method meeting deadline constraint and optimizing cost in cloud
CN117992220B (en) Large model training method based on improved ZeRO-Offload technology
CN111580966B (en) Cloud task scheduling method based on memory utilization rate

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination