CN115914238A - A computing power decomposition method, device and equipment - Google Patents
A computing power decomposition method, device and equipmentDownload PDFInfo
- Publication number
- CN115914238A CN115914238ACN202110890707.4ACN202110890707ACN115914238ACN 115914238 ACN115914238 ACN 115914238ACN 202110890707 ACN202110890707 ACN 202110890707ACN 115914238 ACN115914238 ACN 115914238A
- Authority
- CN
- China
- Prior art keywords
- application
- computing power
- component
- application component
- components
- 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
Links
Images
Classifications
- 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
- Power Sources (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及通信技术领域，特别是指一种算力分解方法、装置及设备。The present invention relates to the field of communication technology, and in particular to a computing power decomposition method, device and equipment.
背景技术Background technique
随着5G网络、边缘计算的规模建设，新兴应用将加速驱动数据处理由云端向边侧、端侧的扩散；同时，云边端的计算能力持续增长，算力泛在化已成趋势。相关技术中，由单一算力节点来承载业务的全部算力需求，如何通过多个算力节点协助来满足业务的算力需求还没有相关方案。With the large-scale construction of 5G networks and edge computing, emerging applications will accelerate the proliferation of data processing from the cloud to the edge and device; at the same time, the computing power of the cloud and edge continues to grow, and the ubiquitous computing power has become a trend. In related technologies, a single computing power node carries all the computing power requirements of the business, and there is no relevant solution on how to meet the computing power requirements of the business through the assistance of multiple computing power nodes.
发明内容Contents of the invention
本发明的目的是提供一种算力分解方法、装置及设备，以解决如何通过多个算力节点协助来满足业务的算力需求的问题。The purpose of the present invention is to provide a computing power decomposition method, device and equipment to solve the problem of how to meet the computing power demand of the business through the assistance of multiple computing power nodes.
为达到上述目的，本发明的实施例提供一种算力分解方法，包括：In order to achieve the above purpose, an embodiment of the present invention provides a computing power decomposition method, including:
根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同；According to the computing power demand information of the application components corresponding to each application mirror in the application mirror package, determine the relationship information of the computing power nodes corresponding to each of the application components, and the computing power nodes corresponding to at least two application mirrors in the application mirror package different;
根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。According to the relationship information of computing power nodes corresponding to each of the application components, the application image in the application image package is repackaged.
可选地，所述算力需求信息包括以下至少一项：Optionally, the computing power demand information includes at least one of the following:
资源需求；resource requirements;
应用组件间的依赖度；Dependencies between application components;
业务约束信息。Business constraint information.
可选地，确定各个所述应用组件对应的算力节点的关系信息，包括：Optionally, determining the relationship information of computing power nodes corresponding to each of the application components includes:
在所述应用镜像包中的第一应用组件对应的第一算力需求信息与第二应用组件对应的第一算力需求信息相同的情况下，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，所述第一算力需求信息包括资源需求和业务约束信息中的至少一项；In the case where the first computing power demand information corresponding to the first application component in the application image package is the same as the first computing power demand information corresponding to the second application component, according to the relationship between the first application component and the second application component Dependency, determine the relationship information of the computing power node corresponding to the first application component and the second application component, the first computing power demand information includes at least one of resource demand and business constraint information;
在第一应用组件对应的第一算力需求信息和第二应用组件对应的第一算力需求信息不同的情况下，确定所述第一应用组件和第二应用组件对应不同的算力节点。In a case where the first computing power requirement information corresponding to the first application component is different from the first computing power requirement information corresponding to the second application component, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，包括：Optionally, according to the dependency between the first application component and the second application component, determining the relationship information of the computing power nodes corresponding to the first application component and the second application component includes:
在第一应用组件和第二应用组件之间的依赖度大于或等于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应同一个算力节点；When the degree of dependence between the first application component and the second application component is greater than or equal to a preset threshold, determining that the first application component and the second application component correspond to the same computing power node;
在第一应用组件和第二应用组件之间的依赖度小于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应不同的算力节点。In a case where the degree of dependence between the first application component and the second application component is less than a preset threshold, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，所述应用组件间的依赖度与以下参数相关：Optionally, the dependency between the application components is related to the following parameters:
应用组件间的数据交互量；The amount of data interaction between application components;
数据依赖度；data dependency;
应用组件间的最大通信时延相关。The maximum communication delay between application components is related.
可选地，根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理，包括：Optionally, according to the relationship information of computing power node information corresponding to each of the application components, repackaging the application image in the application image package includes:
将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包。Package the application images corresponding to the application components corresponding to the same computing power node into an application image package.
可选地，所述根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理之后，还包括：Optionally, after repackaging the application image in the application image package according to the relationship information of the computing power node information corresponding to each of the application components, the method further includes:
对重新打包处理后的应用镜像包进行存储。Store the repackaged application image package.
本发明实施例还提供了一种算力分解装置，包括：The embodiment of the present invention also provides a computing power decomposition device, including:
第一确定模块，用于根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同；The first determination module is configured to determine the relationship information of the computing power nodes corresponding to each application component according to the computing power demand information of the application components corresponding to each application mirror in the application mirroring package, at least two of the application mirroring packages The computing power nodes corresponding to the application image are different;
第一处理模块，用于根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。The first processing module is configured to repackage the application images in the application image package according to the relationship information of the computing power nodes corresponding to each of the application components.
可选地，所述算力需求信息包括以下至少一项：Optionally, the computing power demand information includes at least one of the following:
资源需求；resource requirements;
应用组件间的依赖度；Dependencies between application components;
业务约束信息。Business constraint information.
可选地，所述第一确定模块包括：Optionally, the first determination module includes:
第一确定子模块，用于在所述应用镜像包中的第一应用组件对应的第一算力需求信息与第二应用组件对应的第一算力需求信息相同的情况下，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，所述第一算力需求信息包括资源需求和业务约束信息中的至少一项；The first determination submodule is configured to, in the case that the first computing power requirement information corresponding to the first application component in the application image package is the same as the first computing power requirement information corresponding to the second application component, according to the first application Dependency between the component and the second application component, determine the relationship information of the computing power node corresponding to the first application component and the second application component, and the first computing power demand information includes at least one of resource demand and business constraint information item;
第二确定子模块，用于在第一应用组件对应的第一算力需求信息和第二应用组件对应的第一算力需求信息不同的情况下，确定所述第一应用组件和第二应用组件对应不同的算力节点。The second determining submodule is used to determine the first application component and the second application component in the case that the first computing power demand information corresponding to the first application component is different from the first computing power demand information corresponding to the second application component Components correspond to different computing power nodes.
可选地，所述第一确定子模块用于在第一应用组件和第二应用组件之间的依赖度大于或等于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应同一个算力节点；Optionally, the first determining submodule is configured to determine whether the first application component and the second application component have a dependency between the first application component and the second application component that is greater than or equal to a preset threshold. Application components correspond to the same computing power node;
在第一应用组件和第二应用组件之间的依赖度小于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应不同的算力节点。In a case where the degree of dependence between the first application component and the second application component is less than a preset threshold, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，所述应用组件间的依赖度与以下参数相关：Optionally, the dependency between the application components is related to the following parameters:
应用组件间的数据交互量；The amount of data interaction between application components;
数据依赖度；data dependency;
应用组件间的最大通信时延相关。The maximum communication delay between application components is related.
可选地，所述第一处理模块用于将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包。Optionally, the first processing module is configured to package application images corresponding to application components corresponding to the same computing power node into an application image package.
可选地，本发明实施例的装置，还包括：Optionally, the device of the embodiment of the present invention further includes:
存储模块，用于第一处理模块根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理之后，对重新打包处理后的应用镜像包进行存储。The storage module is used for the first processing module to repackage the application image in the application image package according to the relationship information of the computing power node information corresponding to each of the application components, and then repackage the application image after repackaging The package is stored.
本发明实施例还提供了一种算力分解设备，包括：收发机和处理器；The embodiment of the present invention also provides a computing power decomposition device, including: a transceiver and a processor;
所述处理器用于根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同；根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。The processor is configured to determine, according to the computing power demand information of the application components corresponding to each application image in the application image package, the relationship information of the computing power nodes corresponding to each of the application components, and at least two application images in the application image package The corresponding computing power nodes are different; according to the relationship information of the computing power nodes corresponding to each of the application components, the application image in the application image package is repackaged.
本发明实施例还提供了一种算力分解设备，包括：收发器、处理器、存储器及存储在所述存储器上并可在所述处理器上运行的程序或指令；所述处理器执行所述程序或指令时实现如上所述算力分解方法中的步骤。The embodiment of the present invention also provides a computing power decomposition device, including: a transceiver, a processor, a memory, and a program or instruction stored in the memory and operable on the processor; the processor executes the The above-mentioned steps in the computing power decomposition method are realized when the program or instruction is described.
本发明实施例还提供了一种可读存储介质，其上存储有程序或指令，所述程序或指令被处理器执行时实现如上所述算力分解方法中的步骤。An embodiment of the present invention also provides a readable storage medium, on which a program or an instruction is stored, and when the program or instruction is executed by a processor, the steps in the above computing power decomposition method are implemented.
本发明的上述技术方案的有益效果如下：The beneficial effects of above-mentioned technical scheme of the present invention are as follows:
本发明实施例中，根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，如各个应用组件是否对应同一个算力节点；根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。这样，可将对应同一个算力节点的应用镜像打包成一个应用镜像包，从而实现将原来的应用镜像包对应的全部算力需求分别为重新打包后的各个应用镜像包对应的算力需求，进而能够通过重新打包后的各个应用镜像包对应的多个算力节点协助来满足业务的算力需求。In the embodiment of the present invention, according to the computing power demand information of the application components corresponding to each application image in the application image package, the relationship information of the computing power nodes corresponding to each of the application components is determined, such as whether each application component corresponds to the same computing power A node: repackage the application image in the application image package according to the relationship information of the computing power node corresponding to each of the application components. In this way, the application image corresponding to the same computing power node can be packaged into an application image package, so that all the computing power requirements corresponding to the original application image package can be respectively calculated as the computing power requirements corresponding to each application image package after repackaging. Furthermore, the computing power requirements of the business can be met through the assistance of multiple computing power nodes corresponding to each application mirror package after repackaging.
附图说明Description of drawings
图1为本发明实施例中算力分解方法的流程示意图；Fig. 1 is a schematic flow chart of the computing power decomposition method in the embodiment of the present invention;
图2为本发明实施例中动态算力编排调度系统的结构示意图；Fig. 2 is a schematic structural diagram of a dynamic computing power scheduling system in an embodiment of the present invention;
图3为本发明实施例的算力分解装置的模块示意图；Fig. 3 is a block diagram of a computing power decomposition device according to an embodiment of the present invention;
图4为本发明实施例的算力分解设备的结构框图之一；Fig. 4 is one of the structural block diagrams of the computing power decomposition device according to the embodiment of the present invention;
图5为本发明实施例的算力分解设备的结构框图之二。Fig. 5 is the second structural block diagram of the computing power decomposition device according to the embodiment of the present invention.
具体实施方式Detailed ways
为使本发明要解决的技术问题、技术方案和优点更加清楚，下面将结合附图及具体实施例进行详细描述。In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will describe in detail with reference to the drawings and specific embodiments.
应理解，说明书通篇中提到的“一个实施例”或“一实施例”意味着与实施例有关的特定特征、结构或特性包括在本发明的至少一个实施例中。因此，在整个说明书各处出现的“在一个实施例中”或“在一实施例中”未必一定指相同的实施例。此外，这些特定的特征、结构或特性可以任意适合的方式结合在一个或多个实施例中。It should be understood that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the present invention. Thus, appearances of "in one embodiment" or "in an embodiment" in various places throughout the specification are not necessarily referring to the same embodiment. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner in one or more embodiments.
在本发明的各种实施例中，应理解，下述各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。In various embodiments of the present invention, it should be understood that the sequence numbers of the following processes do not mean the order of execution, and the order of execution of each process should be determined by its functions and internal logic, rather than implementing the present invention. The implementation of the examples constitutes no limitation.
另外，本文中术语“系统”和“网络”在本文中常可互换使用。Additionally, the terms "system" and "network" are often used interchangeably herein.
在本申请所提供的实施例中，应理解，“与A相应的B”表示B与A相关联，根据A可以确定B。但还应理解，根据A确定B并不意味着仅仅根据 A确定B，还可以根据A和/或其它信息确定B。In the embodiments provided in this application, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. But it should also be understood that determining B based on A does not mean determining B based solely on A, and that B can also be determined based on A and/or other information.
如图1所示，本发明实施例提供了一种算力分解方法，该方法包括：As shown in Figure 1, an embodiment of the present invention provides a computing power decomposition method, which includes:
步骤101：根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同。Step 101: According to the computing power demand information of the application components corresponding to each application mirror in the application mirror package, determine the relationship information of the computing power nodes corresponding to each of the application components, and at least two application mirrors in the application mirror package correspond to Computing power nodes are different.
本发明实施例的算力分解方法可基于动态算力编排调度系统实现，如图2 所示，该算力编排调度系统主要包括基础设施层和管理编排层，其中，基础设备层包括网络基础设备与算力基础设施。网络基础设施包括各级软件定义网络控制器(Software DefinedNetwork Controller，SDNC)，网管及相应的网络设备，构建低时延大吞吐、敏捷、无损、确定性算力传输管道；算力基础设施包括运营商自建、合营的核心数据中心(Data Center，DC)与(Mobile Edge Computing，MEC)节点，三方云资源池以及泛在的端算力等，聚焦于异构算力的生成。The computing power decomposition method in the embodiment of the present invention can be realized based on a dynamic computing power scheduling system, as shown in Figure 2, the computing power scheduling system mainly includes an infrastructure layer and a management scheduling layer, wherein the basic equipment layer includes network basic equipment and computing infrastructure. The network infrastructure includes software-defined network controllers (Software Defined Network Controller, SDNC) at all levels, network management and corresponding network equipment, and constructs low-latency, high-throughput, agile, lossless, deterministic computing power transmission pipelines; computing power infrastructure includes operations The core data center (Data Center, DC) and (Mobile Edge Computing, MEC) nodes built by vendors and joint ventures, three-party cloud resource pools, and ubiquitous terminal computing power, etc., focus on the generation of heterogeneous computing power.
管理编排层主要负责算力的注册，算网资源的整合，根据需求进行算力路由调度；算力分解，合理拆分算力需求；部署应用，下发镜像，配置网络、计算资源；安全、策略、资源等管理功能。The management and orchestration layer is mainly responsible for the registration of computing power, the integration of computing network resources, and routing and scheduling of computing power according to requirements; computing power decomposition, reasonably splitting computing power requirements; Management functions such as policies and resources.
本步骤中，各个应用组件对应的算力节点的关系信息包括对应同一个算力节点或对应不同的算力节点。这样，通过确定各个应用组件对应的算力节点的关系信息，从而能够确定是否可以将各个应用组件部署至相同的算力节点。In this step, the relationship information of computing power nodes corresponding to each application component includes corresponding to the same computing power node or corresponding to different computing power nodes. In this way, by determining the relationship information of the computing power nodes corresponding to each application component, it can be determined whether each application component can be deployed to the same computing power node.
上述应用镜像包括内包含一个或多个应用镜像，通常为多个，每个应用镜像对应的组件可视为一个最小调度单元。这里，每个应用镜像可以理解为每个应用镜像对应的应用组件的安装文件。The above application image includes one or more application images, usually multiple, and the components corresponding to each application image can be regarded as a minimum scheduling unit. Here, each application image can be understood as an installation file of an application component corresponding to each application image.
步骤102：根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。Step 102: Repackage the application images in the application image package according to the relationship information of the computing nodes corresponding to each of the application components.
这里，根据每个应用组件对应的算力节点的关系信息可以确定是否将应用组件部署至同一个应用节点，进而可以确定如何对应用镜像包中的应用镜像进行打包处理，如将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包，并生成部署文件和网络配置模型等。Here, according to the relationship information of the computing power nodes corresponding to each application component, it can be determined whether to deploy the application components to the same application node, and then it can be determined how to package the application images in the application image package. The application image corresponding to the application component of the power node is packaged into an application image package, and the deployment file and network configuration model are generated.
本发明实施例中，根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，如各个应用组件是否对应同一个算力节点；根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。这样，可将对应同一个算力节点的应用镜像打包成一个应用镜像包，从而实现将原来的应用镜像包对应的全部算力需求分别为重新打包后的各个应用镜像包对应的算力需求，进而能够通过重新打包后的各个应用镜像包对应的多个算力节点协助来满足业务的算力需求。In the embodiment of the present invention, according to the computing power demand information of the application components corresponding to each application image in the application image package, the relationship information of the computing power nodes corresponding to each of the application components is determined, such as whether each application component corresponds to the same computing power A node: repackage the application image in the application image package according to the relationship information of the computing power node corresponding to each of the application components. In this way, the application image corresponding to the same computing power node can be packaged into an application image package, so that all the computing power requirements corresponding to the original application image package can be respectively calculated as the computing power requirements corresponding to each application image package after repackaging. Furthermore, the computing power requirements of the business can be met through the assistance of multiple computing power nodes corresponding to each application mirror package after repackaging.
可选地，所述算力需求信息包括以下至少一项：Optionally, the computing power demand information includes at least one of the following:
资源需求；resource requirements;
应用组件间的依赖度；Dependencies between application components;
业务约束信息。Business constraint information.
本发明实施例中，上述资源需求可包括需求的计算资源和存储资源。其中，对于计算资源，不同类型的应用组件对算力节点的计算能力要求不同，对计算能力要求较高的应用组件可以部署到具备计算能力强，或具备计算加速资源的节点；对于存储资源，不同类型的应用组件对算力节点的存储能力要求不同，对存储能力要求较高的组件可以部署到存储资源充足的节点。In the embodiment of the present invention, the foregoing resource requirements may include required computing resources and storage resources. Among them, for computing resources, different types of application components have different computing power requirements for computing power nodes. Application components that require high computing power can be deployed to nodes with strong computing power or computing acceleration resources; for storage resources, Different types of application components have different requirements on the storage capacity of computing power nodes, and components with higher storage capacity requirements can be deployed to nodes with sufficient storage resources.
上述应用组件间的依赖度可通过数据交换量和数据依赖度体现。其中，对于数据交互量，不同应用组件间交互会产生通信数据量，对通信性能提出需求。数据交互总量越高，说明两个组件之间具有密切的联系，更应该归入同一个子任务。比如智能安防业务中，视频采集功能组件和预处理原始视频的功能组件会产生较大的数据交互量，这两个功能组件应该放在一起执行。对于数据依赖度，不同应用组件可能读写同一数据块，将读写同一数据块的应用组件分在同一子任务中，能减少物理网络中数据重复传输，降低任务总体的带宽成本。比如智能安防场景中，针对公共摄像头提供的视频流，第一相关部门的视频处理单元会重点关注视频中的人员，而第二相关部门的视频处理单元重点关注车辆，如果两个功能单元放在不同节点独立执行，则需要传输两份相同的视频流给两个部门。而如果这两个功能单元可以归入同一个子任务在同一处节点执行，则视频流数据可以重用，避免网络传输两份相同数据。The dependence among the above application components can be reflected by the amount of data exchange and data dependence. Among them, regarding the amount of data interaction, the interaction between different application components will generate communication data volume, which puts forward requirements on communication performance. The higher the total amount of data interaction, it means that there is a close relationship between the two components, and they should be classified into the same subtask. For example, in the smart security business, the video capture functional component and the functional component for preprocessing the original video will generate a large amount of data interaction, and these two functional components should be executed together. For data dependence, different application components may read and write the same data block, and dividing the application components that read and write the same data block into the same subtask can reduce the repeated transmission of data in the physical network and reduce the overall bandwidth cost of the task. For example, in a smart security scenario, for the video stream provided by a public camera, the video processing unit of the first relevant department will focus on the people in the video, while the video processing unit of the second relevant department will focus on the vehicle. If different nodes are executed independently, two copies of the same video stream need to be transmitted to the two departments. And if these two functional units can be included in the same subtask and executed on the same node, the video stream data can be reused to avoid the network transmission of two copies of the same data.
上述业务约束信息可包括通信时延，该通信时延可包括应用组件处理时延和组件间通信时延，不同业务对时延要求不同。通信时延要求越高的两个应用组件，越应该被归入同一个子任务。比如车联网中，车辆之间、车辆与路边设备之间通过各种无线技术进行连接，需要低延迟聚合点来完成各种协议处理、消息分发和消息处理等工作，路边应用程序需要直接从车辆和路边传感器上的应用程序接收本地信息，分析它们并以非常低的延迟向附近的车辆广播警告 (如事故信息等)。这些应用组件之间尤其需要考虑通信时延。The above service constraint information may include communication delay, which may include application component processing delay and inter-component communication delay, and different services have different delay requirements. Two application components with higher communication latency requirements should be classified into the same subtask. For example, in the Internet of Vehicles, vehicles, and between vehicles and roadside devices are connected through various wireless technologies, requiring low-latency aggregation points to complete various protocol processing, message distribution, and message processing. Roadside applications need to directly Receive local information from applications on vehicles and roadside sensors, analyze them and broadcast warnings (such as accident information, etc.) to nearby vehicles with very low latency. In particular, communication delays need to be considered between these application components.
可选地，确定各个所述应用组件对应的算力节点的关系信息，包括：Optionally, determining the relationship information of computing power nodes corresponding to each of the application components includes:
在所述应用镜像包中的第一应用组件对应的第一算力需求信息与第二应用组件对应的第一算力需求信息相同的情况下，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，所述第一算力需求信息包括资源需求和业务约束信息中的至少一项；In the case where the first computing power demand information corresponding to the first application component in the application image package is the same as the first computing power demand information corresponding to the second application component, according to the relationship between the first application component and the second application component Dependency, determine the relationship information of the computing power node corresponding to the first application component and the second application component, the first computing power demand information includes at least one of resource demand and business constraint information;
在第一应用组件对应的第一算力需求信息和第二应用组件对应的第一算力需求信息不同的情况下，确定所述第一应用组件和第二应用组件对应不同的算力节点。In a case where the first computing power requirement information corresponding to the first application component is different from the first computing power requirement information corresponding to the second application component, it is determined that the first application component and the second application component correspond to different computing power nodes.
本发明实施例中，首先确定两个应用组件对应的资源需求和业务约束信息中的至少一项是否相同，如果两个应用组件对应的资源需求和业务约束信息中的至少一项相同，则根据两个应用组件之间的依赖度进一步确定算力节点的关系信息，否则，确定两个应用组件对应不同的算力节点。In the embodiment of the present invention, it is first determined whether at least one of the resource requirements and business constraint information corresponding to the two application components is the same, and if at least one of the resource requirements and business constraint information corresponding to the two application components is the same, then according to The dependency between the two application components further determines the relationship information of the computing power nodes, otherwise, it is determined that the two application components correspond to different computing power nodes.
例如，如果第一应用组件对应的资源需求和第二应用组件对应的资源需求相同，则根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，如果第一应用组件对应的资源需求和第二应用组件对应的资源需求不同，则确定第一应用组件和第二应用组件对应不同的算力节点。又例如，如果第一应用组件和第二应用组件对应的资源需求和业务约束信息相同，则根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，如果第一应用组件和第二应用组件对应的资源需求和业务约束信息中的至少一者不同，则确定第一应用组件和第二应用组件对应不同的算力节点。For example, if the resource requirements corresponding to the first application component and the resource requirements corresponding to the second application component are the same, then according to the dependency between the first application component and the second application component, it is determined that the first application component and the second application component correspond to If the resource requirements corresponding to the first application component and the resource requirements corresponding to the second application component are different, it is determined that the first application component and the second application component correspond to different computing power nodes. For another example, if the resource requirements and business constraint information corresponding to the first application component and the second application component are the same, then determine the first application component and the second application component according to the dependency between the first application component and the second application component For the relationship information of the corresponding computing power node, if at least one of the resource requirements and business constraint information corresponding to the first application component and the second application component are different, it is determined that the first application component and the second application component correspond to different computing power node.
其中，两个应用组件对应的业务约束信息相同可以理解为两个应用组件对应的通信时延的差值小于预设阈值。The fact that the service constraint information corresponding to the two application components is the same may be understood as that the difference in communication delays corresponding to the two application components is smaller than a preset threshold.
可选地，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，包括：Optionally, according to the dependency between the first application component and the second application component, determining the relationship information of the computing power nodes corresponding to the first application component and the second application component includes:
在第一应用组件和第二应用组件之间的依赖度大于或等于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应同一个算力节点；When the degree of dependence between the first application component and the second application component is greater than or equal to a preset threshold, determining that the first application component and the second application component correspond to the same computing power node;
在第一应用组件和第二应用组件之间的依赖度小于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应不同的算力节点。In a case where the degree of dependence between the first application component and the second application component is less than a preset threshold, it is determined that the first application component and the second application component correspond to different computing power nodes.
上述预设阈值可根据业务场景进行调整，例如，对于应用组件交互敏感的业务场景，如车联网，可增加预设阈值，以保证应用组件尽可能部署在同一算力节点，对于应用组件交互不敏感的业务场景，如密集计算，可通过减少预设阈值的方式，增加应用组件拆解粒度，提升资源利用率。The above preset thresholds can be adjusted according to business scenarios. For example, for business scenarios that are sensitive to application component interaction, such as Internet of Vehicles, the preset threshold can be increased to ensure that application components are deployed on the same computing power node as much as possible. For sensitive business scenarios, such as intensive computing, the granularity of dismantling application components can be increased to improve resource utilization by reducing the preset threshold.
可选地，所述应用组件间的依赖度与以下参数相关：Optionally, the dependency between the application components is related to the following parameters:
应用组件间的数据交互量；The amount of data interaction between application components;
数据依赖度；data dependency;
应用组件间的最大通信时延相关。The maximum communication delay between application components is related.
假设应用组件个数为N，应用组件间的依赖度用矩阵T＝(tij)，应用组件间的数量交互量为C＝(cij)，数据依赖度为D＝(dij)，最大通信时延矩阵为L＝ (lij)，其中，T、C、D、L都是N×N矩阵，tij表示组件i和组件j之间的依赖度，同理，cij、dij、lij分别表示组件i和组件j之间的数据交互量、数据依赖度和最大通信时延，则组件i和组件j间依赖度矩阵与组件间数据交互量、数据依赖度、最大时延的关系可以表示如下：Assuming that the number of application components is N, the dependency between application components is matrix T=(tij ), the quantity interaction between application components is C=(cij ), and the data dependency is D=(dij ), the maximum The communication delay matrix is L= (lij ), where T, C, D, and L are all N×N matrices, and tij represents the dependency between component i and component j. Similarly, cij , dij , lij represent the data interaction amount, data dependence degree and maximum communication delay between component i and component j respectively, then the dependency matrix between component i and component j and the data interaction amount, data dependence degree and maximum delay between components The relationship can be expressed as follows:
tij＝cij*dij/lij。tij=cij *dij/lij .
可选地，根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理，包括：Optionally, according to the relationship information of computing power node information corresponding to each of the application components, repackaging the application image in the application image package includes:
将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包。Package the application images corresponding to the application components corresponding to the same computing power node into an application image package.
这里，根据每个应用组件对应的算力节点的关系信息可以确定是否将应用组件部署至同一个应用节点，进而可以确定如何对应用镜像包中的应用镜像进行打包处理，如将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包，并生成部署文件和网络配置模型等。Here, according to the relationship information of the computing power nodes corresponding to each application component, it can be determined whether to deploy the application components to the same application node, and then it can be determined how to package the application images in the application image package. The application image corresponding to the application component of the power node is packaged into an application image package, and the deployment file and network configuration model are generated.
可选地，所述根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理之后，还包括：Optionally, after repackaging the application image in the application image package according to the relationship information of the computing power node information corresponding to each of the application components, the method further includes:
对重新打包处理后的应用镜像包进行存储。Store the repackaged application image package.
这里，将重新打包处理后的应用镜像包进行存储，如存储至数据库，以便于下次直接应用，而无需再重新分析上述应用镜像包对应的算力节点的关系信息。Here, the repackaged application image package is stored, such as stored in a database, so as to be directly applied next time, without re-analyzing the relationship information of the computing power nodes corresponding to the above application image package.
本发明实施例中，对应用镜像包中应用组件对应的算力需求信息进行分解，将应用组件分开进行部署，输出结果是应用组件的部署方案，应用组件部署后，若没有外界其他条件触发，无需调整。只有在用户位置发生变化、算力节点状态发生变化的情况下，才会触发算随人动、算随人选等动态调整。因此，本发明实施例对算法要求不高，规模较小。In the embodiment of the present invention, the computing power demand information corresponding to the application components in the application image package is decomposed, and the application components are deployed separately. The output result is the deployment plan of the application components. After the application components are deployed, if there are no external conditions triggered, No adjustment is required. Only when the user's location changes and the state of the computing power node changes, will the dynamic adjustments such as calculation follow the person's movement and calculation follow the candidate be triggered. Therefore, the embodiment of the present invention has low requirements on algorithms and a small scale.
本发明的方案具有以下优点：The solution of the present invention has the following advantages:
算力需求分解完成后，算网管理编排平台可以对应用组件进行更小粒度的调度，以提升资源调度的灵活性；After the decomposition of computing power requirements is completed, the computing network management and orchestration platform can schedule application components at a finer granularity to improve the flexibility of resource scheduling;
算力需求分解只与应用需求相关，，不受算力节点状态的影响，因此可以在应用上载时完成，应用部署和调度的过程中无需重新拆解，不会影响部署效率；The decomposition of computing power requirements is only related to application requirements and is not affected by the status of computing power nodes. Therefore, it can be completed when the application is uploaded, and there is no need to re-disassemble the application during deployment and scheduling, and it will not affect deployment efficiency;
算力节点通常存在计算资源和存储资源配比不均匀的情况，导致资源利用率低下。该方案的提出可以将不同算力需求的模块分开部署，提升资源利用效率。Computing power nodes usually have an uneven ratio of computing resources and storage resources, resulting in low resource utilization. The proposal of this solution can deploy modules with different computing power requirements separately to improve resource utilization efficiency.
如图3所示，本发明实施例还提供了一种算力分解装置300，包括：As shown in Figure 3, the embodiment of the present invention also provides a computing
第一确定模块301，用于根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同；The
第一处理模块302，用于根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。The
可选地，所述算力需求信息包括以下至少一项：Optionally, the computing power demand information includes at least one of the following:
资源需求；resource requirements;
应用组件间的依赖度；Dependencies between application components;
业务约束信息。Business constraint information.
可选地，所述第一确定模块包括：Optionally, the first determination module includes:
第一确定子模块，用于在所述应用镜像包中的第一应用组件对应的第一算力需求信息与第二应用组件对应的第一算力需求信息相同的情况下，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，所述第一算力需求信息包括资源需求和业务约束信息中的至少一项；The first determination submodule is configured to, in the case that the first computing power requirement information corresponding to the first application component in the application image package is the same as the first computing power requirement information corresponding to the second application component, according to the first application Dependency between the component and the second application component, determine the relationship information of the computing power node corresponding to the first application component and the second application component, and the first computing power demand information includes at least one of resource demand and business constraint information item;
第二确定子模块，用于在第一应用组件对应的第一算力需求信息和第二应用组件对应的第一算力需求信息不同的情况下，确定所述第一应用组件和第二应用组件对应不同的算力节点。The second determining submodule is used to determine the first application component and the second application component in the case that the first computing power demand information corresponding to the first application component is different from the first computing power demand information corresponding to the second application component Components correspond to different computing power nodes.
可选地，所述第一确定子模块用于在第一应用组件和第二应用组件之间的依赖度大于或等于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应同一个算力节点；Optionally, the first determining submodule is configured to determine whether the first application component and the second application component have a dependency between the first application component and the second application component that is greater than or equal to a preset threshold. Application components correspond to the same computing power node;
在第一应用组件和第二应用组件之间的依赖度小于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应不同的算力节点。In a case where the degree of dependence between the first application component and the second application component is less than a preset threshold, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，所述应用组件间的依赖度与以下参数相关：Optionally, the dependency between the application components is related to the following parameters:
应用组件间的数据交互量；The amount of data interaction between application components;
数据依赖度；data dependency;
应用组件间的最大通信时延相关。The maximum communication delay between application components is related.
可选地，所述第一处理模块用于将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包。Optionally, the first processing module is configured to package application images corresponding to application components corresponding to the same computing power node into an application image package.
可选地，本发明实施例的装置，还包括：Optionally, the device of the embodiment of the present invention further includes:
存储模块，用于第一处理模块根据每个所述应用组件对应的算力节点信息的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理之后，对重新打包处理后的应用镜像包进行存储。The storage module is used for the first processing module to repackage the application image in the application image package according to the relationship information of the computing power node information corresponding to each of the application components, and then repackage the application image after repackaging The package is stored.
该算力分解装置能够实现上述算力分解方法实施例中的所有实现方式，为避免重复，此处不再赘述。The computing power decomposition device can implement all the implementation methods in the foregoing computing power decomposition method embodiments, and to avoid repetition, details are not repeated here.
如图4所示，本发明实施例还提供了一种算力分解设备400，包括：收发机420和处理器410；As shown in Figure 4, the embodiment of the present invention also provides a computing
所述处理器410用于根据应用镜像包中每个应用镜像对应的应用组件的算力需求信息，确定各个所述应用组件对应的算力节点的关系信息，所述应用镜像包中至少两个应用镜像对应的算力节点不同；根据每个所述应用组件对应的算力节点的关系信息，对所述应用镜像包中的应用镜像重新进行打包处理。The
可选地，所述算力需求信息包括以下至少一项：Optionally, the computing power demand information includes at least one of the following:
资源需求；resource requirements;
应用组件间的依赖度；Dependencies between application components;
业务约束信息。Business constraint information.
可选地，所述处理器410还用于：在所述应用镜像包中的第一应用组件对应的第一算力需求信息与第二应用组件对应的第一算力需求信息相同的情况下，根据第一应用组件和第二应用组件之间的依赖度，确定第一应用组件和第二应用组件对应的算力节点的关系信息，所述第一算力需求信息包括资源需求和业务约束信息中的至少一项；Optionally, the
在第一应用组件对应的第一算力需求信息和第二应用组件对应的第一算力需求信息不同的情况下，确定所述第一应用组件和第二应用组件对应不同的算力节点。In a case where the first computing power requirement information corresponding to the first application component is different from the first computing power requirement information corresponding to the second application component, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，所述处理器410还用于：在第一应用组件和第二应用组件之间的依赖度大于或等于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应同一个算力节点；Optionally, the
在第一应用组件和第二应用组件之间的依赖度小于预设阈值的情况下，确定所述第一应用组件和所述第二应用组件对应不同的算力节点。In a case where the degree of dependence between the first application component and the second application component is less than a preset threshold, it is determined that the first application component and the second application component correspond to different computing power nodes.
可选地，所述应用组件间的依赖度与以下参数相关：Optionally, the dependency between the application components is related to the following parameters:
应用组件间的数据交互量；The amount of data interaction between application components;
数据依赖度；data dependency;
应用组件间的最大通信时延相关。The maximum communication delay between application components is related.
可选地，所述处理器410还用于：将对应同一个算力节点的应用组件对应的应用镜像打包为一个应用镜像包。Optionally, the
可选地，所述处理器410还用于：对重新打包处理后的应用镜像包进行存储。Optionally, the
该算力分解设备能够实现上述算力分解方法实施例中的所有实现方式，为避免重复，此处不再赘述。The computing power decomposition device can implement all the implementation methods in the foregoing computing power decomposition method embodiments, and to avoid repetition, details are not repeated here.
本发明一实施例提供了一种数据处理装置，如图5所示，包括收发器510、处理器500、存储器520及存储在所述存储器520上并可在所述处理器500上运行的程序或指令；所述处理器500执行所述程序或指令时实现上述算力分解方法。An embodiment of the present invention provides a data processing apparatus, as shown in FIG. or an instruction; when the
所述收发器510，用于在处理器500的控制下接收和发送数据。The
其中，在图5中，总线架构可以包括任意数量的互联的总线和桥，具体由处理器500代表的一个或多个处理器和存储器520代表的存储器的各种电路链接在一起。总线架构还可以将诸如外围设备、稳压器和功率管理电路等之类的各种其他电路链接在一起，这些都是本领域所公知的，因此，本文不再对其进行进一步描述。总线接口提供接口。收发器510可以是多个元件，即包括发送机和接收机，提供用于在传输介质上与各种其他装置通信的单元。处理器500 负责管理总线架构和通常的处理，存储器520可以存储处理器500在执行操作时所使用的数据。Wherein, in FIG. 5 , the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by the
本发明实施例还提供了一种可读存储介质，其上存储有程序或指令，所述程序或指令被处理器执行时实现如上所述的算力分解方法中的步骤，且能达到相同的技术效果，为避免重复，这里不再赘述。The embodiment of the present invention also provides a readable storage medium on which programs or instructions are stored. When the programs or instructions are executed by a processor, the steps in the above computing power decomposition method can be realized, and the same Technical effects, in order to avoid repetition, will not be repeated here.
其中，所述处理器为上述实施例中所述的算力分解装置中的处理器。所述可读存储介质，包括计算机可读存储介质，如计算机只读存储器(Read-Only Memory，简称ROM)、随机存取存储器(Random Access Memory，简称RAM)、磁碟或者光盘等。Wherein, the processor is the processor in the computing power decomposition device described in the above embodiment. The readable storage medium includes computer readable storage medium, such as computer read-only memory (Read-Only Memory, ROM for short), random access memory (Random Access Memory, RAM for short), magnetic disk or optical disk, etc.
进一步需要说明的是，此说明书中所描述的终端包括但不限于智能手机、平板电脑等，且所描述的许多功能部件都被称为模块，以便更加特别地强调其实现方式的独立性。It should be further noted that the terminals described in this manual include but are not limited to smartphones, tablet computers, etc., and many of the described functional components are called modules, in order to more particularly emphasize the independence of their implementation.
本发明实施例中，模块可以用软件实现，以便由各种类型的处理器执行。举例来说，一个标识的可执行代码模块可以包括计算机指令的一个或多个物理或者逻辑块，举例来说，其可以被构建为对象、过程或函数。尽管如此，所标识模块的可执行代码无需物理地位于一起，而是可以包括存储在不同位里上的不同的指令，当这些指令逻辑上结合在一起时，其构成模块并且实现该模块的规定目的。In the embodiments of the present invention, the modules may be implemented by software so as to be executed by various types of processors. An identified module of executable code may, by way of example, comprise one or more physical or logical blocks of computer instructions which may, for example, be structured as an object, procedure, or function. Notwithstanding, the executable code of an identified module need not be physically located together, but may include distinct instructions stored in different bits which, when logically combined, constitute the module and implement the specified Purpose.
实际上，可执行代码模块可以是单条指令或者是许多条指令，并且甚至可以分布在多个不同的代码段上，分布在不同程序当中，以及跨越多个存储器设备分布。同样地，操作数据可以在模块内被识别，并且可以依照任何适当的形式实现并且被组织在任何适当类型的数据结构内。所述操作数据可以作为单个数据集被收集，或者可以分布在不同位置上(包括在不同存储设备上)，并且至少部分地可以仅作为电子信号存在于系统或网络上。Indeed, a module of executable code may be a single instruction, or many instructions, and may even be distributed over several different code segments, among different programs and across multiple memory devices. Likewise, operational data may be identified within modules, and may be implemented in any suitable form and organized within any suitable type of data structure. The operational data may be collected as a single data set, or may be distributed in different locations (including on different storage devices), and may exist, at least in part, only as electronic signals on a system or network.
在模块可以利用软件实现时，考虑到现有硬件工艺的水平，所以可以以软件实现的模块，在不考虑成本的情况下，本领域技术人员都可以搭建对应的硬件电路来实现对应的功能，所述硬件电路包括常规的超大规模集成(VLSI) 电路或者门阵列以及诸如逻辑芯片、晶体管之类的现有半导体或者是其它分立的元件。模块还可以用可编程硬件设备，诸如现场可编程门阵列、可编程阵列逻辑、可编程逻辑设备等实现。When the module can be realized by software, considering the level of the existing hardware technology, the module that can be realized by software, regardless of the cost, those skilled in the art can build the corresponding hardware circuit to realize the corresponding function. The hardware circuit includes conventional very large scale integration (VLSI) circuits or gate arrays as well as existing semiconductors such as logic chips, transistors, or other discrete components. A module may also be implemented in programmable hardware devices such as field programmable gate arrays, programmable array logic, programmable logic devices, and the like.
上述范例性实施例是参考该些附图来描述的，许多不同的形式和实施例是可行而不偏离本发明精神及教示，因此，本发明不应被建构成为在此所提出范例性实施例的限制。更确切地说，这些范例性实施例被提供以使得本发明会是完善又完整，且会将本发明范围传达给那些熟知此项技术的人士。在该些图式中，组件尺寸及相对尺寸也许基于清晰起见而被夸大。在此所使用的术语只是基于描述特定范例性实施例目的，并无意成为限制用。如在此所使用地，除非该内文清楚地另有所指，否则该单数形式“一”、“一个”和“该”是意欲将该些多个形式也纳入。会进一步了解到该些术语“包含”及/或“包括”在使用于本说明书时，表示所述特征、整数、步骤、操作、构件及/或组件的存在，但不排除一或更多其它特征、整数、步骤、操作、构件、组件及/或其族群的存在或增加。除非另有所示，陈述时，一值范围包含该范围的上下限及其间的任何子范围。The exemplary embodiments described above are described with reference to these drawings. Many different forms and embodiments are possible without departing from the spirit and teachings of the present invention. Therefore, the present invention should not be construed as the exemplary embodiments set forth herein. limits. Rather, these exemplary embodiments are provided so that this disclosure will be thorough and complete, and will convey the scope of the invention to those skilled in the art. In the drawings, component sizes and relative sizes may be exaggerated for clarity. The terminology used herein is for the purpose of describing certain exemplary embodiments only and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include these plural forms unless the context clearly dictates otherwise. It will be further understood that the terms "comprises" and/or "comprises", when used in this specification, indicate the presence of stated features, integers, steps, operations, components and/or components, but do not exclude one or more other The presence or addition of features, integers, steps, operations, components, components and/or groups thereof. Unless otherwise indicated, when stated a range of values includes the upper and lower limits of that range and any subranges therebetween.
以上所述是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明所述原理的前提下，还可以作出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。The above description is a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.
Claims (17)
Translated fromChinesePriority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110890707.4ACN115914238A (en) | 2021-08-04 | 2021-08-04 | A computing power decomposition method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110890707.4ACN115914238A (en) | 2021-08-04 | 2021-08-04 | A computing power decomposition method, device and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115914238Atrue CN115914238A (en) | 2023-04-04 |
Family
ID=86490062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110890707.4APendingCN115914238A (en) | 2021-08-04 | 2021-08-04 | A computing power decomposition method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115914238A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190138287A1 (en)* | 2017-11-09 | 2019-05-09 | International Business Machines Corporation | Dynamic selection of deployment configurations of software applications |
| CN111800442A (en)* | 2019-04-08 | 2020-10-20 | 阿里巴巴集团控股有限公司 | Network system, mirror image management method, device and storage medium |
| CN112084006A (en)* | 2020-09-10 | 2020-12-15 | 杭州海康威视系统技术有限公司 | Mirror image packet processing method and device and electronic equipment |
| CN112256282A (en)* | 2020-09-24 | 2021-01-22 | 深圳云天励飞技术股份有限公司 | Application mirror image construction method and device, electronic equipment and storage medium |
| CN113037786A (en)* | 2019-12-09 | 2021-06-25 | 中国电信股份有限公司 | Intelligent computing power scheduling method, device and system |
- 2021
- 2021-08-04CNCN202110890707.4Apatent/CN115914238A/enactivePending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190138287A1 (en)* | 2017-11-09 | 2019-05-09 | International Business Machines Corporation | Dynamic selection of deployment configurations of software applications |
| CN111800442A (en)* | 2019-04-08 | 2020-10-20 | 阿里巴巴集团控股有限公司 | Network system, mirror image management method, device and storage medium |
| CN113037786A (en)* | 2019-12-09 | 2021-06-25 | 中国电信股份有限公司 | Intelligent computing power scheduling method, device and system |
| CN112084006A (en)* | 2020-09-10 | 2020-12-15 | 杭州海康威视系统技术有限公司 | Mirror image packet processing method and device and electronic equipment |
| CN112256282A (en)* | 2020-09-24 | 2021-01-22 | 深圳云天励飞技术股份有限公司 | Application mirror image construction method and device, electronic equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 张智雄: "面向大规模异构评测任务的容器集群管理方法", 中国优秀硕士学位论文全文数据库信息科技辑, 15 January 2021 (2021-01-15)* |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xiao et al. | Distributed optimization for energy-efficient fog computing in the tactile internet | |
| Mseddi et al. | Joint container placement and task provisioning in dynamic fog computing | |
| Zhang et al. | Firework: Data processing and sharing for hybrid cloud-edge analytics | |
| CN107087019B (en) | A task scheduling method and device based on a terminal-cloud collaborative computing architecture | |
| EP3734453A1 (en) | Modular i/o configurations for edge computing using disaggregated chiplets | |
| WO2021190482A1 (en) | Computing power processing network system and computing power processing method | |
| WO2020168761A1 (en) | Model training method and apparatus | |
| CN115278639A (en) | Perceptual data acquisition method, device, device and storage medium | |
| CN115328663A (en) | Method, device, equipment and storage medium for scheduling resources based on PaaS platform | |
| CN110430068A (en) | Characteristic engineering arrangement method and device | |
| CN112667400A (en) | Edge cloud resource scheduling method, device and system managed and controlled by edge autonomous center | |
| Miao et al. | Performance modelling and quantitative analysis of vehicular edge computing with bursty task arrivals | |
| CN115002215B (en) | Resource allocation model training method and resource allocation method for government and enterprises on the cloud | |
| CN113364888B (en) | Service scheduling method, system, electronic device and computer readable storage medium | |
| Phung et al. | OneVFC—A vehicular fog computation platform for artificial intelligence in Internet of Vehicles | |
| WO2024011908A1 (en) | Network prediction system and method, and electronic device and storage medium | |
| Rath et al. | MAQ system development in mobile ad-hoc networks using mobile agents | |
| Nasser et al. | An efficient Time-sensitive data scheduling approach for Wireless Sensor Networks in smart cities | |
| CN110247783A (en) | A kind of scalable appearance policy conflict processing method and processing device | |
| WO2023093200A1 (en) | Method and apparatus for asynchronously processing tasks, and storage medium and electronic apparatus | |
| Ye et al. | EdgeStreaming: Secure computation intelligence in distributed edge networks for streaming analytics | |
| US20240126606A1 (en) | Dynamic parallel processing in an edge computing system | |
| CN115914238A (en) | A computing power decomposition method, device and equipment | |
| CN119127469A (en) | Computing resource allocation method, device, computer equipment, readable storage medium and program product | |
| Chen et al. | Topology-aware Microservice Architecture in Edge Networks: Deployment Optimization and Implementation |
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 |