WO2023284830A1

WO2023284830A1 - Management and scheduling method and apparatus, node, and storage medium

Info

Publication number: WO2023284830A1
Application number: PCT/CN2022/105717
Authority: WO
Inventors: 姚惠娟; 付月霞; 陆璐; 孙滔
Original assignee: 中国移动通信有限公司研究院; 中国移动通信集团有限公司
Priority date: 2021-07-14
Filing date: 2022-07-14
Publication date: 2023-01-19
Also published as: CN115622904A

Abstract

The present application discloses a management and scheduling method and apparatus, a node, and a storage medium. The method comprises: a first node managing a computing power resource and a network resource of a network; managing a computing power service of the network; and receiving a service request of a first service, and scheduling the first service.

Description

Management and scheduling method, device, node and storage medium

Cross References to Related Applications

This application is based on a Chinese patent application with application number 202110797438.7 and a filing date of July 14, 2021, and claims the priority of this Chinese patent application. The entire content of this Chinese patent application is hereby incorporated by reference into this application.

technical field

The present application relates to the field of data communication, and in particular to a management and scheduling method, device, node and storage medium.

Background technique

Under the general trend of cloud computing and edge computing development, in the future society, many computing power of different scales will be distributed at different distances close to users, and these computing power will provide users with various personalized services through the global network. From tens of billions of smart terminals, to billions of global home gateways, to thousands of edge clouds with computing capabilities brought by future multi-access edge computing (MEC) in each city, and every Dozens of large-scale cloud data centers (DC) in the country can form massive ubiquitous computing power to connect to the Internet from everywhere, thus forming a development trend of deep integration of computing (that is, computing power) and the network.

For the deep integration of computing and network, as shown in Figure 1, by integrating computing resources in the network into every corner of the network, each network node can become a resource provider; and user requests can be made by calling the nearest To meet the needs of node resources, and no longer limited to a specific node, to avoid the waste of connection and network scheduling resources. Compared with traditional networks, traditional networks only provide channels for data communication, are based on connections, and are subject to fixed network addressing mechanisms. They often cannot meet user needs under higher and more stringent quality of experience (QoE) requirements. In addition, with the development of microservices, the traditional network client (English can be expressed as client)-server (English can be expressed as server) mode is deconstructed. The application on the server side is deconstructed into functional components and deployed on the cloud platform, and is uniformly scheduled by the application program interface (API) gateway (English can be expressed as Gateway), which can realize on-demand dynamic instantiation. The business logic on the server side is transferred to the client side, and the client only needs to care about the computing function itself, and does not need to care about computing resources such as servers, virtual machines, containers, etc., thereby realizing Function as a Service (FaaS, Function as a Service).

It can be seen that under the development trend of future-oriented network computing and network deep integration, based on the dynamic and distributed computing realized by network nodes, users can directly obtain the calculation results through the network without caring about computing resources. Enable the network to meet the user's QoE requirements.

However, how to optimize the network architecture to improve the rationality of service distribution is still a technical problem to be solved urgently.

Contents of the invention

To solve related technical problems, embodiments of the present application provide a management and scheduling method, device, node, and storage medium.

The technical scheme of the embodiment of the application is realized in this way:

The embodiment of this application provides a management and scheduling method applied to the first node, including:

Manage network computing resources and network resources;

Manage computing power services of the network;

The service request of the first service is received, and the first service is scheduled.

In the above solution, the scheduling of the first service includes:

generating a scheduling policy for the first service based at least on the computing resources and network resources of the network;

and sending the scheduling policy to the second node of the network; the scheduling policy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service into the network The corresponding third node performs processing; the second node has at least a network control function; and the third node has at least a computing power sensing function and a forwarding function.

In the above solution, the scheduling of the first service includes:

sending the computing power resource information of the network to the second node of the network, and the sent computing power resource information is used for the second node to generate the computing power resource information for the first service based on at least the computing power resource information and network resources a scheduling strategy; the scheduling strategy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service to a corresponding third node in the network for processing; the second The node has at least a network control function and a computing power information sensing function; the third node has at least a computing power sensing function and a forwarding function.

In the above solution, the scheduling of the first service includes:

Sending the computing power resource information and network resource information of the network to at least one third node of the network, the sent computing power resource information and network resource information are used for the third node based on at least the computing power resource information and the network resource information Resource information, generating a scheduling policy for the first service; the scheduling policy is used for the third node to determine the forwarding path of the first service, so as to schedule the first service to a corresponding The third node for processing; the third node has at least a computing power sensing function and a forwarding function.

In the above scheme, the method also includes:

Obtain the status information of the computing resources of the network;

The computing resources of the network are managed based on the acquired status information of the computing resources.

In the above scheme,

Updating the computing power resource topology information of the network based on the obtained computing power resource state information.

In the above scheme, the method also includes:

Obtaining network resource status information of the network;

The network resources of the network are managed based on the acquired network resource status information.

In the above scheme,

Updating the network resource topology information of the network based on the acquired network resource state information.

In the above solution, when managing the computing resources and network resources of the network, the method includes:

Perform Operation Administration and Maintenance (OAM, Operation Administration and Maintenance) operations on the computing resources and network resources of the network;

and / or,

Perform operational processing on the computing resources and/or network resources of the network.

In the above solution, the management of the computing service of the network includes at least one of the following:

Manage computing power service images;

Manage instances of computing services;

Manage the resources corresponding to the computing power service.

In the above solution, the management of the computing power service image includes at least one of the following:

Added new computing power service image;

Update the version of the computing power service image;

Delete the computing service image.

In the above solution, the management of computing power service instances includes at least one of the following:

Notifying the second node to establish a connection between nodes and a connection between nodes and terminals;

Evaluate the service quality of the computing power service based at least on the computing power resource status information of the network, the network resource status information and the computing power service status information;

Dispatch the computing power service to at least one fourth node to perform at least one of the following operations:

Instantiate the computing power service;

Update the instance of computing power service;

Termination of computing services;

Wherein, the fourth node has at least a computing power function.

In the above solution, the management of the resources corresponding to the computing service includes at least one of the following:

Reserve resources for computing services;

Allocate resources for computing services;

Release resources for computing services;

Expand and/or shrink the resources of the computing power service.

The embodiment of the present application also provides a management and scheduling device, including:

The first management unit is configured to manage computing resources and network resources of the network;

The second management unit is configured to manage the computing service of the network;

The scheduling unit is configured to receive the service request of the first service, and schedule the first service.

The embodiment of the present application also provides a node, including: a processor and a communication interface; wherein,

The processor is configured to manage network computing resources and network resources; manage network computing services; and receive a service request for a first service through the communication interface, and schedule the first service .

The embodiment of the present application also provides a node, including: a processor and a memory configured to store a computer program that can run on the processor,

Wherein, the processor is configured to execute the steps of any one of the above methods when running the computer program.

The embodiment of the present application also provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of any one of the above methods are implemented.

The management and scheduling method, device, node, and storage medium provided by the embodiments of the present application manage network computing power resources and network resources; manage network computing power services; receive a service request for the first business, and The first service is scheduled. In the solution of the embodiment of the present application, the first node is introduced into the network architecture, and the computing power resources, network resources and computing power services of the network are managed through the first node, and the business is scheduled through the first node; in this way, it is possible to realize Unified management of network computing resources, network resources and computing services, and flexible scheduling of services, so that the network architecture can meet the needs of computing and network integration and evolution, and enable reasonable distribution of services, thereby improving user experience.

Description of drawings

Figure 1 is a schematic diagram of the development trend of computing and network deep integration in related technologies;

FIG. 2 is a schematic diagram of the architecture of a computing power-aware network (CAN, Computing-aware Networking) according to an embodiment of the present application;

FIG. 3 is a schematic flow diagram of a management and scheduling method in an embodiment of the present application;

Fig. 4 is the architecture schematic diagram of the application embodiment CAN of the present application;

Fig. 5 is a schematic diagram of the network architecture of the application embodiment CAN;

Fig. 6 is a schematic diagram of a computing network collaborative orchestration management in an application embodiment of the present application;

FIG. 7 is a schematic diagram of another computing-network collaborative orchestration management in the application embodiment of the present application;

Fig. 8 is a schematic diagram of the third computing-network collaborative orchestration management of the application embodiment of the present application;

FIG. 9 is a schematic structural diagram of a management and scheduling device according to an embodiment of the present application;

FIG. 10 is a schematic diagram of the node structure of the embodiment of the present application;

Fig. 11 is a schematic structural diagram of the management and scheduling system of the embodiment of the present application.

detailed description

The application will be further described in detail below in conjunction with the accompanying drawings and embodiments.

The new-generation network architecture design for the future network needs to consider the needs of network and computing convergence and evolution, and realize global optimization of the network in the ubiquitous connection and computing power architecture, flexible scheduling of computing power, and reasonable distribution of services.

Based on this, the embodiment of the present application provides a CAN architecture, as shown in Figure 2, the CAN architecture includes: a first processing layer, a second processing layer, a third processing layer, a fourth processing layer and a fifth processing layer; wherein The first processing layer is configured to carry various services and applications of ubiquitous computing; the second processing layer is configured to comprehensively consider the status of network resources and computing power resources, and flexibly schedule services to corresponding nodes on demand; The third processing layer is configured to support computing power registration, computing power operation, computing power notification and other functions; the fourth processing layer is configured to use various computing infrastructure to provide computing power resources; the fifth processing layer is configured to Utilize various network infrastructures to provide ubiquitous network connections for every corner of the network.

The CAN architecture in the embodiment of this application can interconnect dynamically distributed computing resources based on ubiquitous network connections, and through the unified collaborative scheduling of multi-dimensional resources such as network, storage, and computing power, massive applications can be invoked on demand and in real time Computing resources in different places realize the global optimization of connection and computing power in the network and provide a consistent user experience. In actual application, CAN can also be called a future data communication network, a computing network, a computing power network, a computing power endogenous network, or a computing-network integration network, etc., as long as the functions of the network are realized, the name of the network in the embodiment of this application Not limited.

It should be noted that the processing layer in the embodiment of the present application is a virtual layer structure divided according to logical functions. In actual deployment, the above-mentioned processing layers can be deployed on one device or on multiple devices; if deployed On one device, each processing layer can transmit information through internal interfaces; if it is deployed on multiple devices, each processing layer can realize information transmission through signaling interaction.

In actual application, the first processing layer can also be called computing power application layer or computing power service layer; the second processing layer can also be called routing layer or computing power routing layer; the third processing layer It can also be called the computing network management and arrangement layer, the computing network arrangement layer, the computing network orchestration management layer, the computing power platform layer, the computing power management platform layer or the computing power management layer, etc.; the fourth processing layer can also be called computing power resource layer, etc.; the fifth processing layer may also be called a network resource layer; the embodiment of the present application does not limit the names of each processing layer, as long as the functions of each processing layer can be realized.

Based on the above-mentioned CAN architecture, in various embodiments of the present application, a node is introduced into the CAN architecture, through which the computing power resources, network resources and computing power services of the network are managed, and services are scheduled through the node , that is, through this node, the collaborative orchestration and management of the computing network can be realized, and the unified management of computing resources, network resources, and computing services of the network can be realized, and flexible scheduling of services can be realized, so that the network architecture can meet the requirements of computing and network convergence evolution. demand, and enable reasonable distribution of services, which in turn can improve user experience.

The embodiment of the present application provides a management and scheduling method, which is applied to the first node, as shown in FIG. 3, the method includes:

Step 301: Manage network computing resources and network resources;

Step 302: Manage the computing service of the network;

Step 303: Receive the service request of the first service, and schedule the first service.

Here, in actual application, step 301, step 302 and step 303 are executed in no particular order.

In practical applications, the first node may also be called a computing network orchestration center, a computing power orchestration management center, a computing power scheduling and scheduling center, a computing network scheduling center, or a computing network unified scheduling center. The name of the first node is not limited, as long as the function of the first node can be realized.

In practical applications, the network refers to a future data communication network that is deeply integrated with computing and network, which may be called CAN, or a computing power network, etc.

In practical applications, the computing resources may include computing resources of nodes with computing capabilities in the network, and the nodes may be understood as network devices. For example, the computing resources may include computing resources of processors such as a single-core central processing unit (CPU) or a multi-core CPU in a network device; for another example, the computing resources may include CPUs, graphics processing units (GPUs) ), Field-Programmable Gate Array (FPGA, Field-Programmable Gate Array) and other combinations of computing resources of at least two processors.

In practical applications, the computing resources may also include storage resources (expressed as Storage in English) of nodes with computing capabilities in the network. For example, the computing power resources may also include storage resources such as random access memory (RAM, Random Access Memory) or read-only memory (ROM, Read Only Memory) in the network device; A combination of storage resources of at least two types of memory such as RAM and ROM in the network device may be included.

In practical applications, the nodes with computing capabilities in the network may include a third node and a fourth node; the third node may at least have a computing power sensing function and a forwarding function, such as a router with a computing power sensing function; The fourth node can at least have a computing power function, such as a DC server. Here, nodes with computing capabilities in the network may also be referred to as computing power network element nodes, etc. The embodiment of the present application does not limit the names of such nodes, as long as they have computing capabilities.

In practical applications, the third node may also be called a computing power routing node, etc. The embodiment of the present application does not limit the name of the third node, as long as the function of the third node can be realized.

In practical applications, the fourth node may also be called a computing power node, etc. The embodiment of the present application does not limit the name of the fourth node, as long as the function of the fourth node can be realized.

In practical applications, the network resources may include network resources of networks such as access networks, metropolitan area networks, and backbone networks, such as bandwidth, delay, and jitter.

In practical application, the network needs to be controlled, therefore, the network may further include a second node having a network control function.

In practical applications, the second node may also be called a network controller or a computing network controller, etc. The embodiment of the present application does not limit the name of the second node, as long as the functions of the second node can be realized .

In step 301, in actual application, the first node needs to have a computing power awareness function, so that when the first node manages the computing power resources of the network, it can perceive (that is, obtain) each computing power resource status of a node with computing power, and manage the computing power resources of the network according to the obtained computing power resource status.

Based on this, in an embodiment, the method may also include:

Obtain the status information of the computing resources of the network;

Correspondingly, the management of network computing resources may include:

In practical application, the first node may acquire the status information of computing power resources of the third node and the fourth node of the network.

Wherein, the computing resource status information of the third node may include at least one of the following:

An identifier corresponding to the third node (such as a service identification number (ID));

The device type corresponding to the third node;

The chip type corresponding to the third node;

The number of service links supported by the third node;

Computing resource information of the third node; such as CPU, GPU, FPGA and other processor status information;

Storage resource information of the third node; such as status information of storage such as memory and hard disk.

In practical applications, the first node may acquire the computing power resource status information of the third node in the network in any of the following three ways:

Mode 1: the first node notifies (that is, notifies) all third nodes in the network, and each third node directly reports the state information of the computing resources to the first node;

Method 2: The first node notifies all third nodes in the network, and each third node reports the status information of computing resources to the second node or the first processing layer, and the second node or The first processing layer reports the received computing resource status information to the first node;

Mode 3, the first node notifies the second node or the first processing layer to report the computing power resource status information of the third node, and the second node or the first processing layer obtains computing power resources from the third node status information, and send the acquired computing power resource status information of the third node to the first node.

Here, in the above method 1, method 2 and method 3, in order to improve the efficiency of managing the computing power resources of the network, the computing power resource status information of the third node can be generated based on the specified template, that is, the first node sends The third node issues a template for reporting the state of the computing power resource, and the third node reports the state information of the computing power resource based on the received template. In this way, after the first node obtains the status information of the computing resources, it can shield (that is, ignore) the differences of the underlying hardware devices, so as to realize efficient management of the computing resources of the network.

Based on this, in an embodiment, the acquiring the status information of the computing resources of the network may include:

sending at least one first template to the third node;

receiving computing resource status information reported by the third node based on the at least one first template.

Wherein, in actual application, the first template is used to abstractly describe and represent the state of the computing resources of the third node, so that the first node can realize efficient management of heterogeneous computing resources. Here, heterogeneous computing resources can be understood as heterogeneous computing resources and/or storage resources, which means that there are differences between two network devices with computing capabilities at the hardware level, for example, the CPU, GPU, and bus interface of a network device The model of processing hardware such as chip (BIC, Bus Interface Chip), digital signal processor (DSP, Digital Signal Processor) and/or storage hardware such as RAM and ROM is different from that of another network device.

In actual application, when the first node sends the first template to the third node, it may send a unified first template to all third nodes, or send a unified first template to third nodes belonging to a specific type (For example, at least two third nodes supporting the same type of business) issue a unified first template, and can also issue a specific first template for a specific third node; in other words, each third node receives At least one first template can be the same or different.

In actual application, the first node may directly send the first template to the third node, or send the first template to the third node through the second node or the first processing layer. template; correspondingly, the computing resource status information reported by the third node based on the at least one first template may be directly sent to the first node, or sent through the second node or the first processing layer to the first node.

In practical applications, when the first node sends at least one first template to the third node, it may also instruct the third node to report the frequency of computing resource status information, and the specific method of determining the frequency can be set according to requirements , which is not limited in this embodiment of the present application. In addition, the indication information of the frequency may be included in the first template, or the first node may separately send the indication information of the frequency to the third node.

In actual application, any one of the above three methods can be selected according to requirements to enable the first node to obtain the computing power resource status information of the third node, or other methods can be adopted according to requirements to enable the first node to obtain the third node. The computing power resource status information of the node. The embodiment of this application does not limit the specific method for the first node to obtain the computing power resource status information of the third node, as long as the first node can obtain the computing power of the third node Resource status information is sufficient.

In practical applications, the computing resource status information of the fourth node may include at least one of the following:

An identifier corresponding to the fourth node (such as a service ID);

The device type corresponding to the fourth node;

The chip type corresponding to the fourth node;

The number of service links supported by the fourth node;

Computing resource information of the fourth node; such as CPU, GPU, FPGA and other processor status information;

Storage resource information of the fourth node; such as status information of memory, hard disk and other storage.

In practical applications, the first node may acquire the computing resource status information of the fourth node in the network in any of the following three ways:

Mode 1: the first node notifies all fourth nodes in the network, and each fourth node directly reports the status information of computing resources to the first node;

Mode 2: The first node notifies all fourth nodes in the network, and each fourth node reports the status information of computing resources to the second node or the first processing layer, and the second node or The first processing layer reports the received computing resource status information to the first node;

Mode 3, the first node notifies the second node or the first processing layer to report the computing resource status information of the fourth node, and the second node or the first processing layer obtains the computing resource from the fourth node state information, and send the obtained state information of computing power resources of the fourth node to the first node.

Here, in the above method 1, method 2 and method 3, in order to improve the efficiency of managing the computing power resources of the network, the computing power resource status information of the fourth node can be generated based on the specified template, that is, the first node sends The fourth node issues a template for reporting the state of the computing power resource, and the fourth node reports the state information of the computing power resource based on the received template. In this way, after the first node obtains the status information of the computing resources, it can shield (that is, ignore) the differences of the underlying hardware devices, so as to realize efficient management of the computing resources of the network.

sending at least one second template to the fourth node;

receiving computing resource status information reported by the fourth node based on the at least one second template.

Wherein, in actual application, the second template is used to abstractly describe and represent the state of the computing resources of the fourth node, so that the first node can realize efficient management of heterogeneous computing resources.

In actual application, when the first node sends the second template to the fourth node, it may send a unified second template to all fourth nodes, or send a unified second template to fourth nodes belonging to a specific type (For example, at least two fourth nodes supporting the same type of business) issue a unified second template, and can also issue a specific second template for a specific fourth node; in other words, each fourth node receives At least one second template can be the same or different.

In practical applications, when the first node issues a template for reporting computing power resource status information to nodes with computing capabilities in the network, it is not necessary to distinguish the types of nodes. In other words, the first node The templates delivered to the third node and the fourth node may be the same or different.

In actual application, the first node may directly send the second template to the fourth node, or send the second template to the fourth node through the second node or the first processing layer. template; correspondingly, the computing resource status information reported by the fourth node based on the at least one second template may be directly sent to the first node, or sent through the second node or the first processing layer to the first node.

In actual application, when the first node sends at least one second template to the fourth node, it may also instruct the fourth node to report the frequency of computing resource status information, and the specific method of determining the frequency can be set according to requirements , which is not limited in this embodiment of the present application. In addition, the indication information of the frequency may be included in the second template, or the first node may separately send the indication information of the frequency to the fourth node.

In actual application, any one of the above three methods can be selected according to requirements to enable the first node to obtain the computing power resource status information of the fourth node, or other methods can be used according to requirements to enable the first node to obtain the fourth node. The computing power resource status information of the node, the embodiment of the present application does not limit the specific method for the first node to obtain the computing power resource status information of the fourth node, as long as the first node can obtain the computing power of the fourth node Resource status information is sufficient.

In practical applications, the status information of the computing power resources changes in real time and is a kind of dynamic information; therefore, in order to further improve the efficiency of managing the computing power resources of the network, the first node may obtain The state information of computing power resources of the third node and/or the fourth node is obtained, and static and global network computing power resource topology information is generated, and the computing power resource topology information of the network is calculated by using the generated computing power resource topology information. Human resources are managed. In addition, when acquiring new computing power resource state information, the first node may update the computing power resource topology information to ensure the timeliness of the computing power resource topology information.

Based on this, in an embodiment, the method may also include:

Correspondingly, in actual application, the management of the computing resources of the network may include:

The updated computing resource topology information is used to manage the computing resources of the network.

In practical applications, the first node manages the computing power resources of the network, which may specifically include registering, updating and deregistering the third node and/or the fourth node. For example, after the third node and/or the fourth node go online, it may send computing power resource status information to the first node, and the first node The status information of computing resources of the node is registered with the third node and/or the fourth node.

In practical application, the first node also needs to perceive the state of the network, so as to manage the network resources of the network.

Based on this, in an embodiment, the method may also include:

Obtaining network resource status information of the network;

Correspondingly, the managing network resources of the network may include:

The network resources of the network are managed based on the acquired network resource state information.

In practical applications, the third node may also have a network awareness function, that is, the first node may obtain network resource status information of the third node, and the network resource status information may include network bandwidth, delay, Status information such as delay jitter.

In actual application, the first node can obtain the network resource status information of the third node in any of the following three ways:

Mode 1: the first node notifies all third nodes in the network, and each third node directly reports network resource status information to the first node;

Mode 2: the first node notifies all third nodes in the network, each third node reports the network resource status information to the second node, and the second node reports the received network resource status information to the first node;

Mode 3: The first node notifies the second node to report the network resource status information of the third node, the second node obtains the network resource status information from the third node, and sends the acquired network resource status information of the third node to to the first node.

Here, in the above method 1, method 2 and method 3, in order to improve the efficiency of managing the network resources of the network, the network resource status information of the third node can be generated based on a specified template, that is, the first node sends the third node The node issues a template for reporting the status of the network resource, and the third node reports the status information of the network resource based on the received template. In this way, after the first node obtains the network resource state information, it can shield (ie ignore) the hardware difference between the network infrastructures, so as to realize efficient management of the network resources of the network.

Based on this, in an embodiment, the acquiring the network resource state information of the network may include:

sending at least one third template to the third node;

receiving network resource status information reported by the third node based on the at least one third template.

In practical application, the third template is used to abstractly describe and represent the state of the network resources perceived by the third node, so that the first node can realize efficient management of heterogeneous network resources. Here, heterogeneous network resources can be understood as differences between two network infrastructures at the hardware level.

In actual application, when the first node sends the third template to the third node, it may send a unified third template to all third nodes, or send a unified third template to third nodes belonging to a specific type (For example, at least two third nodes supporting the same type of business) issue a unified third template, and can also issue a specific third template for a specific third node; in other words, each third node receives At least one third template can be the same or different.

In actual application, the first node may directly send the third template to the third node, or send the third template to the third node through the second node; correspondingly, the The network resource state information reported by the third node based on the at least one third template may be directly sent to the first node, or sent to the first node through the second node.

In practical applications, when the first node sends at least one third template to the third node, it may also instruct the third node to report the frequency of network resource status information, and the specific method of determining the frequency may be set according to requirements, This embodiment of the present application does not limit it. In addition, the indication information of the frequency may be included in the third template, or the first node may separately send the indication information of the frequency to the third node.

In practical applications, the frequency at which the third node reports network resource status information and the frequency at which the third node reports computing power resource status information may be the same or different, which is not limited in this embodiment of the present application.

In actual application, any one of the above three methods can be selected to enable the first node to obtain network resource status information according to requirements, or other methods can be used to enable the first node to obtain network resource status information. The embodiment does not limit the specific manner in which the first node obtains the network resource state information, as long as the first node can obtain the network resource state information of the network.

In actual application, the network resource status information changes in real time and is a kind of dynamic information; therefore, in order to further improve the efficiency of managing the network resources of the network, the first node may, according to the obtained network The resource state information generates static and global network resource topology information, and uses the generated network resource topology information to manage the network resources of the network. Moreover, when acquiring new network resource state information, the first node may update the network resource topology information, so as to ensure the timeliness of the network resource topology information.

Based on this, in an embodiment, the method may also include:

Correspondingly, in actual application, the managing the network resources of the network may include:

The network resources of the network are managed by using the updated network resource topology information.

In practical applications, when the first node manages the computing power resources and network resources of the network, it also needs to maintain and operate the computing power resources and network resources.

Based on this, in an embodiment, when managing the computing resources and network resources of the network, the method may further include:

Perform OAM operations on the computing resources and network resources of the network;

and / or,

In actual application, the OAM operation may include performance monitoring, fault detection and alarm, link loop test, etc.; the operation processing may include user computing power service contract, billing management, etc.

In step 302, in actual application, the first node may obtain computing power service status information of the network, and manage computing power service of the network based on the obtained computing power service status information.

In actual application, the computing power service may also be called application service, or directly referred to as service; it can be understood as a description of the application on the server side. Exemplarily, the computing service may include virtual reality (VR), augmented reality (AR), vehicle wireless communication technology (V2X), Internet of Things (IoT) and/or application programs (APP). The type of computing power service can be set according to requirements, which is not limited in this embodiment of the present application.

In actual application, the first node may specifically obtain the computing power service status information of the fourth node; the computing power service status information may include computing power service image related information, computing power service corresponding resource related information, Information about the instantiation of the computing power service, etc.

In practical applications, the manner in which the first node obtains the computing power service status information of the fourth node can be set according to requirements, for example, the first node can obtain the fourth node's service status information through the first processing layer Computing power service status information; for another example, the first node may obtain computing power service status information of the fourth node through the second node.

In actual application, in order to improve the efficiency of managing the computing power service of the network, the computing power service status information of the fourth node can be generated based on a specified template, that is, the first node sends the computing power to the fourth node for reporting A service status template, the fourth node reports computing power service status information based on the received template. In this way, after the first node obtains the computing power service status information, it can shield (that is, ignore) the difference of the underlying hardware devices, so as to realize efficient management of the computing power service of the network.

Based on this, in actual application, the first node may send at least one fourth template to the fourth node, and receive computing power service status information reported by the fourth node based on the at least one fourth template.

Here, the fourth template is used to abstractly describe and represent the computing service status of the fourth node.

In actual application, when the first node issues the fourth template to the fourth node, it may issue a unified fourth template for all fourth nodes, or issue a fourth template for a specific type of fourth node (For example, at least two fourth nodes that support the same type of business) issue a unified fourth template, and can also issue a specific fourth template for a specific fourth node; in other words, each fourth node receives At least one fourth template can be the same or different.

In actual application, the first node may directly send the fourth template to the fourth node, or send the fourth template to the fourth node through the second node, or send the fourth template to the fourth node through the The first processing layer sends the fourth template to the fourth node; correspondingly, the computing power service status information reported by the fourth node based on the at least one fourth template may be directly sent to the first node, Either send to the first node through the second node, or send to the first node through the first processing layer.

In practical applications, when the first node sends at least one fourth template to the fourth node, it may also instruct the fourth node to report the frequency of computing power service status information, and the specific method of determining the frequency can be set according to requirements , which is not limited in this embodiment of the present application. In addition, the indication information of the frequency may be included in the fourth template, or the first node may separately send the indication information of the frequency to the fourth node.

In practical applications, the frequency at which the fourth node reports computing power service status information and the frequency at which the fourth node reports computing power resource status information may be the same or different, which is not limited in this embodiment of the present application.

In practical applications, the computing power service status information changes in real time and is a kind of dynamic information; therefore, in order to further improve the efficiency of managing the computing power service of the network, the first node can obtain The computing power service status information of the computing power service, generate static, global network computing power service topology information, and use the generated computing power service topology information to manage the computing power service of the network. Moreover, when acquiring new computing power service status information, the first node may update the computing power service topology information to ensure the timeliness of the computing power service topology information.

Based on this, in actual application, the first node can update the computing power service topology information of the network based on the obtained computing power service status information; correspondingly, the first node can use the updated computing power service topology information , to manage the computing service of the network.

For step 302, in an embodiment, the management of the computing service of the network may include at least one of the following:

Manage computing power service images;

Manage instances of computing services;

Manage the resources corresponding to the computing power service.

Here, the first node may directly implement the management of the computing service, or may implement the management of the computing service through the first processing layer, which is not limited in this embodiment of the present application.

In practical applications, the management of computing power services can be realized based on information communication between nodes and information communication between nodes and the processing layer.

In practical applications, the management of the computing service of the network may also include other management related to the lifecycle of the computing service, which is not limited in this embodiment of the present application.

Wherein, in an embodiment, the management of the computing power service image may include at least one of the following:

Added new computing power service image;

Update the version of the computing power service image;

Delete the computing service image.

In an embodiment, the managing of computing power service instances may include at least one of the following:

Instantiate the computing power service;

Update the instance of computing power service;

Terminate computing service.

In practical application, the management of instances of computing power services can be understood as orchestrating computing power services.

In actual application, the notification to the second node to establish the connection between nodes and the connection between nodes and the terminal can be understood as the end-to-end realization of the computing power service to the terminal, that is, the establishment of the connection between the fourth node and the terminal. The communication link between the nodes; the connection between the nodes may include the connection between the third node and the fourth node; the connection between the nodes and the terminal may include the connection between the third node and the terminal.

In practical applications, after the connection between nodes and between nodes and terminals is established, a service level agreement (SLA, Service Level Agreement) of the corresponding level can be provided according to the quality requirements of the computing power service.

In practical application, the computing power service can be dispatched to at least one fourth node based on a preset strategy or a preset artificial intelligence (AI) algorithm (such as a machine learning model trained using historical data in advance).

In practical applications, the updating of instances of computing power services may include adding or deleting instances of computing power services.

In practical applications, the termination of the computing service may include terminating the instance of the computing service.

In an embodiment, the management of the resources corresponding to the computing power service may include at least one of the following:

Reserve resources for computing services;

Allocate resources for computing services;

Release resources for computing services;

Expand and/or shrink the resources of the computing power service.

Here, the resources may include computing power resources of the network.

In step 303, in actual application, the service request can be understood as the user's demand for the first service, that is, the demand that the first service itself needs to be satisfied when implementing the first service, such as bandwidth demand, Latency requirements, Quality of Service (QoS), etc.

In practical application, the first service may be supported by at least one computing power service.

In step 303, in actual application, the first node needs to comprehensively consider service requirements, computing power resource status information, and network resource status information to generate a coordinated scheduling policy to schedule the first service. Specifically, the first node may schedule the first service in any of the following three ways:

Mode 1: The first node performs scheduling on the management plane, that is, the first node directly generates a scheduling policy for the first service, and sends the scheduling policy to the second node, and the second node The second node schedules the first service according to the scheduling policy.

Based on this, in an embodiment, the scheduling of the first service may include:

and sending the scheduling strategy to the second node; the scheduling strategy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service to a corresponding The third node is processed.

Mode 2: The first node performs scheduling on the control plane, that is, the first node sends the computing power resource information of the network to the second node, and the second node generates the information of the first service A scheduling strategy, and scheduling the first service according to the generated scheduling strategy.

sending computing power resource information of the network to the second node, and the sent computing power resource information is used for the second node to generate a scheduling strategy for the first service based on at least computing power resource information and network resources ; The scheduling policy is used for the second node to determine a forwarding path of the first service, so as to schedule the first service to a corresponding third node in the network for processing.

In practical applications, the computing power resource information of the network may be used to reflect the global static computing power resources of the network, such as the computing power resource topology information.

In practical applications, since the second node has a network control function, the second node can determine the network resource status of the network in real time, in other words, after receiving the computing resource information, the The second node may directly generate the scheduling policy based on the computing power resource information and its own network resource information.

Mode 3: The first node performs scheduling on the data plane, that is, the first node sends the computing resource information and network resource information of the network to the third node, and the third node generates the a scheduling strategy for the first service, and schedule the first service according to the generated scheduling strategy.

sending the computing power resource information and network resource information of the network to at least one of the third nodes, and the sent computing power resource information and network resource information are used for the third node based on at least the computing power resource information and the network resource information , generating a scheduling policy for the first service; the scheduling policy is used for the third node to determine a forwarding path of the first service, so as to schedule the first service to a corresponding first service in the network Three nodes are processed.

In practical applications, the computing power resource information of the network can be used to reflect the global static computing power resources of the network, such as the topology information of the computing power resources; the network resource information of the network can be used to reflect the network Global static network resources, such as topology information of the network resources.

In actual application, any one of the above three methods can be selected to schedule the first service according to requirements, or other methods can be used to schedule the first service according to requirements. The specific scheduling manner of a service is not limited, as long as the first service can be scheduled to the corresponding third node for processing.

The management and scheduling method provided by the embodiment of the present application manages network computing power resources and network resources; manages network computing power services; receives a service request of a first business, and schedules the first business. In the solution of the embodiment of the present application, the first node is introduced into the network architecture, and the computing power resources, network resources and computing power services of the network are managed through the first node, and the business is scheduled through the first node; in this way, it is possible to realize Unified management of network computing resources, network resources and computing services, and flexible scheduling of services, so that the network architecture can meet the needs of computing and network integration and evolution, and enable reasonable distribution of services, thereby improving user experience.

The present application will be further described in detail below in conjunction with application examples.

In this application example, as shown in Figure 4, the CAN architecture can be logically divided into five functional modules: computing power service layer, computing network management and arrangement layer, computing power resource layer, computing power routing layer and network resource layer , can support computing-network collaborative orchestration and management, and realize unified operation and maintenance management of computing power resources and network resources. Through service-oriented network reconstruction, computing-aware network routing, and computing-network collaborative orchestration management, a brand new The information and communication technology (ICT) infrastructure makes the network the infrastructure of a smart society.

Among them, the computing power service layer is configured to carry various services and applications of ubiquitous computing, and supports distributed micro-service architecture, that is, it supports decomposing applications into atomic functional components and forming algorithm libraries, which are uniformly scheduled by API Gateway , to realize functions such as service decomposition and service scheduling.

The computing power routing layer includes a control plane and a forwarding plane; the computing power routing layer is configured to discover computing network resources based on abstraction (that is, the computing power resource status information generated based on the first template or the second template and based on The network resource status information generated by the third template), comprehensively consider the network status and computing power status, and flexibly schedule services (such as the above-mentioned first service) to different computing resource nodes (that is, the above-mentioned third node and/or the first four nodes).

The computing network management and orchestration layer is configured to support the registration, update and registration of computing power nodes (that is, the above-mentioned fourth node), network nodes (that is, the above-mentioned third node) and service information (that is, the above-mentioned computing power service) of the entire network. Cancellation and other management, that is, to support the registration of the computing power resource layer, network resource layer and computing power service layer with the computing power dispatching and orchestration center (also called the computing network orchestration management center or computing network orchestration and scheduling center) to generate computing power, services and Network topology information.

The computing power resource layer is configured to use computing infrastructure to provide computing power resources, and in order to meet the diverse computing needs of the edge computing field, it is oriented to different applications and provides computing power models and computing power APIs on the basis of physical computing resources , Computing network resource identification and other functions. Among them, the computing infrastructure can include a combination of various computing capabilities ranging from single-core CPU, multi-core CPU to CPU+GPU+FPGA.

The network resource layer is configured to use network infrastructure to provide ubiquitous network connections for every corner of the network; wherein, the network infrastructure may include an access network, a metropolitan area network, and a backbone network.

In this application example, the computing power resource layer and the network resource layer are the infrastructure layers of the CAN architecture; the computing network management and arrangement layer and the computing power routing layer are two core functional modules of the CAN architecture , users and applications access CAN through the computing power routing layer, and realize perception, control and scheduling of computing power resources and network resources through the computing network management and orchestration layer.

In this application embodiment, the computing network management and orchestration layer may include subfunctional modules such as the computing network orchestration management center, computing power resource management center, and network management center.

Wherein, as shown in FIG. 5 , the computing network orchestration management center is configured to complete unified management of computing power resources and network resources, including perception, measurement and OAM management of computing power resources and network resources. Specifically, the computing network orchestration management center can perceive computing power resources and network resources, construct global topology information of computing power (that is, the above computing power resource topology information) and network global topology information (that is, the above network resource topology information), and construct Service global topology information (that is, the above computing power service topology information), based on the constructed global topology information, realizes the unified operation of computing power resources and network resources. At the same time, the computing network orchestration management center can also be configured based on the received computing power information (that is, the above-mentioned computing power resource status information), network information (that is, the above-mentioned network resource status information) and service information (that is, the above-mentioned computing power service status information), and generate a coordinated scheduling strategy for computing power resources and network resources according to business requirements (that is, the above-mentioned service requests).

The computing power resource management center is first configured to abstractly describe and represent computing power resources through computing power modeling for heterogeneous computing resources to form node computing power information (that is, the above-mentioned first template or second template) , to shield the difference of the underlying hardware devices; the computing power information can be transmitted to the corresponding network node through the computing power notification (such as the CAN routing node shown in Figure 5, the function of the CAN routing node is equivalent to the function of the third node above) . Secondly, it can also be configured to perform OAM operations on computing power resources and network resources, and realize computing power operations and network operations. Here, the computing power resource management center needs to receive the configuration and management of the computing network orchestration management center, and report the computing power status (that is, the above computing power resource status information) to the computing network orchestration management center.

The network management center is configured to implement management and operation and maintenance of current network resources. Here, the network management center needs to receive the configuration and management of the computing network orchestration management center, and report the network resource status (that is, the above-mentioned network resource status information) to the computing network orchestration management center.

In this application embodiment, the computing network orchestration management center can also be configured to support the management of registration, update and cancellation of computing power nodes, network nodes and service information of the entire network. For example, after a computing power node goes online, it can notify its computing power enabling information to the computing network orchestration management center (that is, send the above computing power resource status information to the computing network orchestrating management center for the first time), and the computing power enabling information can include Resource information such as power node identification or computing power resource identification, device type, chip type, storage, and computing.

In this application embodiment, the computing network orchestration management center can also be configured to realize the configuration and management of the method of perception of services, network resources and computing power resources, mainly including but not limited to:

1) Send multi-dimensional resource-aware templates (such as the above-mentioned first template, second template, third template and fourth template) to the computing power service layer, computing power resource layer and network resource layer for the computing power service layer, The computing power resource layer and the network resource layer perform computing power service information (that is, the above computing power service status information), computing power resource information (that is, the above computing power resource status information) and network resource information (that is, the above network resource status information) according to the template. perception and measurement.

2) Through the multi-dimensional resource-aware template or separately delivered instruction information, the computing network orchestration management center can adaptively order the parameters that need to be collected, sensed or measured, and the frequency of feedback (that is, the frequency of reporting information), etc. . According to the received perception and measurement results of multi-dimensional resources reported according to the specified frequency (including the above-mentioned computing power resource status information, the above-mentioned network resource status information and the above-mentioned computing power service status information), the computing network management and arrangement layer can update the computing power topology information , service topology information, and network topology information, and then based on the updated computing power topology information, service topology information, and network topology information, the collaborative orchestration of network resources and computing power can be programmed and the automatic adaptation of services can be realized.

3) It can receive end-user requirements (that is, the above-mentioned service requests), comprehensively consider business requirements, computing resource status information and network resource status information, and generate a collaborative scheduling strategy to achieve flexible scheduling of services.

4) Support the management function of computing power service, that is, support the life cycle management related capabilities of computing power service on each computing power node, which may specifically include but not limited to:

Manage the computing power service image, including adding, version updating, deleting, etc.;

Instantiation, update, expansion, shrinkage, instance termination, etc. of computing power services;

Reserve, allocate and release heterogeneous computing power resources (such as CPU, GPU, embedded neural network processor (NPU), FPGA, etc.) on computing power nodes.

5) Support computing power service orchestration function, that is, for computing power service experience, intelligently arrange and dispatch computing power services to their suitable computing power nodes, which may include but not limited to:

Based on the preset strategy or AI algorithm, it interacts with the computing power service management function (that is, the computing power service layer) to realize the instantiation of the computing power service on one or more computing power nodes according to the computing power service quality requirements, Service update, flexible expansion and contraction, service termination, etc.;

Notify computing power routing nodes (such as software-defined network (SDN) network controllers) to establish connections between nodes and nodes and between nodes and users, and provide corresponding levels of SLA according to computing power service quality requirements;

Through the perception of computing power node status, network status, and computing power service status, the evaluation of computing power service quality is realized.

In actual application, the computing power service mainly describes the application from the server side; the computing network management and orchestration layer can directly manage and arrange the computing power service, or through the perception of computing power node status, network status, and computing power service status , generate the computing power service scheduling strategy and send it to the computing power service layer, and the computing power service layer realizes the management and arrangement of computing power services according to the received strategy; in other words, the decomposition, execution and scheduling of computing power services can be It can be executed at the power service layer, or at the computing network management and orchestration layer.

In this application embodiment, according to the network deployment status, the computing network management and orchestration layer can be selected on the management plane (that is, the computing network orchestration management center, corresponding to the above-mentioned first node), the control plane (that is, the computing power The control plane of the routing layer, corresponding to the above-mentioned second node) or the data plane (that is, the computing power routing node, corresponding to the above-mentioned third node) executes specific scheduling.

The three computing-network collaborative scheduling methods will be described in detail below with reference to FIG. 6 to FIG. 8 .

First, as shown in Figure 6, the computing network collaborative scheduling is performed on the management plane, that is, the "network management module" of the computing network orchestration management center notifies the "computing power orchestrator + network (also called computing network orchestrator)" Network information (that is, the above-mentioned network resource status information), the computing network orchestration management center performs a unified computing network collaborative scheduling, generates a scheduling strategy, and sends the scheduling strategy to the network controller (also called the computing network controller). The network controller further generates a path forwarding table according to the scheduling strategy, so that by enhancing the interface configuration between the "computing power orchestrator" and the network controller, that is, the network controller sends network information to the computing power orchestrator, and the computing power orchestration The server perceives network resource information and improves the efficiency of executing collaborative scheduling strategies.

Wherein, the network controller is configured to collect network information, report the network information to the computing network orchestrator, and is configured to receive the network orchestration policy (ie, the scheduling policy) of the computing network orchestrator. The computing network orchestrator is configured to collect computing power information (that is, the above computing power resource status information), receive network information from the network controller, perform joint programming of computing power resources and network resources, and generate an orchestration strategy; and is configured to Send the orchestration policy to the network controller. It can be understood that the computing network orchestrator is responsible for service scheduling.

Secondly, as shown in Figure 7, the computing network collaborative scheduling is performed on the control plane, that is, the "computing power orchestrator" of the computing network orchestration management center notifies the computing power information to the computing network controller through the "network management module", and the computing network The controller performs coordinated scheduling of unified computing power resources and network resources, generates scheduling policies, and further generates path forwarding tables according to the scheduling policies, thereby enhancing the interface configuration between the "computing power orchestrator" and the network controller, and The network controller is enhanced to allow the network controller to perceive computing resource information and improve the efficiency of executing collaborative scheduling strategies.

Third, as shown in Figure 8, the computing network collaborative scheduling is performed on the data plane, that is, the "computing network unified orchestrator" of the computing network orchestration management center performs collaborative scheduling of the network and computing power, and is constructed by the computing network management and orchestration layer Static computing power topology information and network topology information, the computing power topology information and network topology information are delivered to the data plane, and the data plane realizes the generation and execution of collaborative scheduling strategies; in other words, the computing network management and orchestration layer Topology information and network topology information are delivered to the control plane, which implements distributed service scheduling.

In this application example, each module of the Computing Network Orchestration Management Center is divided according to logical functions. In actual application, each function of the Computing Network Orchestration Management Center can be divided into different modules according to requirements.

The solution provided by this application example has the following advantages:

First, through the collaborative orchestration and management of the computing network, the unified control and management of computing power resources and network resources can be realized, and the global optimal configuration of computing power resources and network resources can be realized.

Second, provide a variety of computing-network collaborative orchestration and management methods in which the management plane (that is, the computing network orchestration management center), the control plane (that is, the computing network controller) and the data plane (that is, the computing power routing node) cooperate The solution is conducive to phased deployment on the live network and helps to promote the practical progress of the computing power network.

In order to implement the method of the embodiment of the present application, the embodiment of the present application also provides a management and scheduling device, as shown in Figure 9, the device includes:

The first management unit 901 is configured to manage computing resources and network resources of the network;

The second management unit 902 is configured to manage computing power services of the network;

The scheduling unit 903 is configured to receive the service request of the first service, and schedule the first service.

Wherein, in an embodiment, the scheduling unit 903 is further configured to:

In an embodiment, the scheduling unit 903 is further configured to send computing power resource information of the network to the second node of the network, and the sent computing power resource information is used for the second node based on at least computing resource information and network resources, and generate a scheduling policy for the first service; the scheduling policy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service to the The corresponding third node in the network performs processing; the second node has at least a network control function and a computing power information sensing function; the third node has at least a computing power sensing function and a forwarding function.

In an embodiment, the scheduling unit 903 is further configured to send computing power resource information and network resource information of the network to at least one third node of the network, and the sent computing power resource information and network resource information are used in For the third node to generate a scheduling strategy for the first service based on at least computing power resource information and network resource information; the scheduling strategy is used for the third node to determine the forwarding path of the first service , so as to dispatch the first service to a corresponding third node in the network for processing; the third node has at least a computing power sensing function and a forwarding function.

In an embodiment, the device further includes an acquiring unit configured to acquire status information of computing power resources of the network; the third node has at least a computing power sensing function and a forwarding function;

Correspondingly, the first management unit 901 is further configured to manage the computing resources of the network based on the acquired status information of the computing resources.

In an embodiment, the device further includes an updating unit configured to update the computing power resource topology information of the network based on the obtained computing power resource state information.

In an embodiment, the obtaining unit is further configured to obtain network resource state information of the network;

Correspondingly, the first management unit 901 is further configured to manage the network resources of the network based on the acquired network resource state information.

In an embodiment, the updating unit is further configured to update the network resource topology information of the network based on the acquired network resource status information.

In an embodiment, when the first management unit 901 manages the computing resources and network resources of the network, it is further configured to:

and / or,

In an embodiment, the second management unit 902 is further configured to perform one of the following operations:

Manage computing power service images;

Manage instances of computing services;

Manage the resources corresponding to the computing power service.

Wherein, in an embodiment, the second management unit 902 is further configured to perform one of the following operations:

Added new computing power service image;

Update the version of the computing power service image;

Delete the computing service image.

Instantiate the computing power service;

Update the instance of computing power service;

Termination of computing services;

Wherein, the fourth node has at least a computing power function.

Reserve resources for computing services;

Allocate resources for computing services;

Release resources for computing services;

Expand and/or shrink the resources of the computing power service.

In actual application, the first management unit 901, the second management unit 902, the scheduling unit 903, the acquiring unit and the updating unit may be implemented by a processor in the management and scheduling device.

It should be noted that: when the management and scheduling device provided by the above-mentioned embodiment schedules services, it only uses the division of the above-mentioned program modules for illustration. In actual application, the above-mentioned processing can be assigned to different program modules according to needs Completion means that the internal structure of the device is divided into different program modules to complete all or part of the processing described above. In addition, the management and scheduling device and the management and scheduling method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process thereof is detailed in the method embodiments, and will not be repeated here.

Based on the hardware implementation of the above program modules, and in order to implement the method of the embodiment of the present application, the embodiment of the present application also provides a node, as shown in FIG. 10 , the node 1000 includes:

Communication interface 1001, capable of information exchange with other nodes;

The processor 1002 is connected to the communication interface 1001 to implement information interaction with other nodes, and is configured to execute the methods provided by one or more of the above technical solutions when running a computer program;

The memory 1003 stores computer programs that can run on the processor 1002 .

Specifically, the processor 1002 is configured to:

Manage network computing resources and network resources;

Manage computing power services of the network;

Wherein, in an embodiment, the processor 1002 is further configured to:

Generate a scheduling strategy for the first service based at least on the computing resources and network resources of the network;

In an embodiment, the processor 1002 is further configured to send computing power resource information of the network to a second node of the network, and the sent computing power resource information is used by the second node based on at least computing resource information and network resources, and generate a scheduling policy for the first service; the scheduling policy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service to the The corresponding third node in the network performs processing; the second node has at least a network control function and a computing power information sensing function; the third node has at least a computing power sensing function and a forwarding function.

In an embodiment, the processor 1002 is further configured to send computing power resource information and network resource information of the network to at least one third node of the network, and the sent computing power resource information and network resource information are used in For the third node to generate a scheduling strategy for the first service based on at least computing power resource information and network resource information; the scheduling strategy is used for the third node to determine the forwarding path of the first service , so as to dispatch the first service to a corresponding third node in the network for processing; the third node has at least a computing power sensing function and a forwarding function.

In an embodiment, the processor 1002 is further configured to:

Obtaining computing power resource status information of the network; the third node has at least a computing power sensing function and a forwarding function;

In an embodiment, the processor 1002 is further configured to update the computing power resource topology information of the network based on the obtained computing power resource state information.

In an embodiment, the processor 1002 is further configured to:

Obtaining network resource status information of the network;

In an embodiment, the processor 1002 is further configured to update the network resource topology information of the network based on the acquired network resource state information.

In an embodiment, when the processor 1002 manages the computing resources and network resources of the network, it is further configured to:

and / or,

In an embodiment, the processor 1002 is further configured to perform one of the following operations:

Manage computing power service images;

Manage instances of computing services;

Manage the resources corresponding to the computing power service.

Wherein, in an embodiment, the processor 1002 is further configured to perform one of the following operations:

Added new computing power service image;

Update the version of the computing power service image;

Delete the computing service image.

Instantiate the computing power service;

Update the instance of computing power service;

Termination of computing services;

Wherein, the fourth node has at least a computing power function.

Reserve resources for computing services;

Allocate resources for computing services;

Release resources for computing services;

Expand and/or shrink the resources of the computing power service.

It should be noted that: the specific process of the processor 1002 performing the above operations can be found in the method embodiments, and will not be repeated here.

Certainly, in actual application, various components in the node 1000 are coupled together through the bus system 1004 . It can be understood that the bus system 1004 is used to realize connection and communication between these components. In addition to the data bus, the bus system 1004 also includes a power bus, a control bus and a status signal bus. However, the various buses are labeled as bus system 1004 in FIG. 10 for clarity of illustration.

The memory 1003 in the embodiment of the present application is used to store various types of data to support the operation of the node 1000 . Examples of such data include: any computer program for operating on node 1000 .

The methods disclosed in the foregoing embodiments of the present application may be applied to the processor 1002 or implemented by the processor 1002 . The processor 1002 may be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above-mentioned method may be completed by an integrated logic circuit of hardware in the processor 1002 or instructions in the form of software. The aforementioned processor 1002 may be a general-purpose processor, DSP, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. The processor 1002 may implement or execute various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the method disclosed in the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, and the storage medium is located in the memory 1003, and the processor 1002 reads the information in the memory 1003, and completes the steps of the aforementioned method in combination with its hardware.

In an exemplary embodiment, the node 1000 may be implemented by one or more Application Specific Integrated Circuit (ASIC, Application Specific Integrated Circuit), DSP, Programmable Logic Device (PLD, Programmable Logic Device), Complex Programmable Logic Device (CPLD, Complex Programmable Logic Device), FPGA, general-purpose processor, controller, microcontroller (MCU, Micro Controller Unit), microprocessor (Microprocessor), or other electronic components are used to implement the aforementioned method.

It can be understood that the memory 1003 in this embodiment of the present application may be a volatile memory or a nonvolatile memory, and may also include both volatile and nonvolatile memories. Among them, the non-volatile memory can be ROM, Programmable Read-Only Memory (PROM, Programmable Read-Only Memory), Erasable Programmable Read-Only Memory (EPROM, Erasable Programmable Read-Only Memory), Electrically Erasable Programmable read-only memory (EEPROM, Electrically Erasable Programmable Read-Only Memory), magnetic random access memory (FRAM, ferromagnetic random access memory), flash memory (Flash Memory), magnetic surface memory, optical disc, or CD-ROM -ROM, Compact Disc Read-Only Memory); magnetic surface storage can be disk storage or tape storage. Volatile memory can be RAM, which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM, Static Random Access Memory), Synchronous Static Random Access Memory (SSRAM, Synchronous Static Random Access Memory), Dynamic Random Access Memory Memory (DRAM, Dynamic Random Access Memory), synchronous dynamic random access memory (SDRAM, Synchronous Dynamic Random Access Memory), double data rate synchronous dynamic random access memory (DDRSDRAM, Double Data Rate Synchronous Dynamic Random Access Memory), enhanced Synchronous Dynamic Random Access Memory (ESDRAM, Enhanced Synchronous Dynamic Random Access Memory), Synchronous Link Dynamic Random Access Memory (SLDRAM, SyncLink Dynamic Random Access Memory), Direct Memory Bus Random Access Memory (DRRAM, Direct Rambus Random Access Memory ). The memories described in the embodiments of the present application are intended to include but not limited to these and any other suitable types of memories.

In order to implement the method provided by the embodiment of the present application, the embodiment of the present application also provides a management and scheduling system, as shown in Figure 11, the system includes: a first node 1101, a second node 1102, a third node 1103 and Four nodes 1104 .

Here, it should be noted that: the specific processing procedures of the first node 1101, the second node 1102, the third node 1103, and the fourth node 1104 have been described in detail above, and will not be repeated here.

In an exemplary embodiment, the embodiment of the present application also provides a storage medium, that is, a computer storage medium, specifically a computer-readable storage medium, for example, including a memory 1003 storing a computer program, and the above-mentioned computer program can be executed by the processor of the node 1000 Step 1002 is executed to complete the steps described in the foregoing method. The computer-readable storage medium can be memories such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

It should be noted that: "first", "second", etc. are used to distinguish similar objects, and not necessarily used to describe a specific order or sequence.

In addition, the technical solutions described in the embodiments of the present application may be combined arbitrarily if there is no conflict.

The above descriptions are only preferred embodiments of the present application, and are not intended to limit the protection scope of the present application.

Claims

A management and scheduling method, applied to a first node, comprising:

Manage network computing resources and network resources;

Manage computing power services of the network;

The service request of the first service is received, and the first service is scheduled.
The method according to claim 1, wherein the scheduling of the first service comprises:

generating a scheduling policy for the first service based at least on the computing resources and network resources of the network;

and sending the scheduling policy to the second node of the network; the scheduling policy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service into the network The corresponding third node performs processing; the second node has at least a network control function; and the third node has at least a computing power sensing function and a forwarding function.
The method according to claim 1, wherein the scheduling of the first service comprises:

sending the computing power resource information of the network to the second node of the network, and the sent computing power resource information is used for the second node to generate the computing power resource information for the first service based on at least the computing power resource information and network resources a scheduling strategy; the scheduling strategy is used for the second node to determine the forwarding path of the first service, so as to schedule the first service to a corresponding third node in the network for processing; the second The node has at least a network control function and a computing power information sensing function; the third node has at least a computing power sensing function and a forwarding function.
The method according to claim 1, wherein the scheduling of the first service comprises:

Sending the computing power resource information and network resource information of the network to at least one third node of the network, the sent computing power resource information and network resource information are used for the third node based on at least the computing power resource information and the network resource information Resource information, generating a scheduling policy for the first service; the scheduling policy is used for the third node to determine the forwarding path of the first service, so as to schedule the first service to a corresponding The third node for processing; the third node has at least a computing power sensing function and a forwarding function.
The method according to claim 1, wherein the method further comprises:

Obtain the status information of the computing resources of the network;

The computing resources of the network are managed based on the acquired status information of the computing resources.
The method according to claim 5, wherein,

Updating the computing power resource topology information of the network based on the obtained computing power resource state information.
The method according to claim 1, wherein the method further comprises:

Obtaining network resource status information of the network;

The network resources of the network are managed based on the acquired network resource state information.
The method according to claim 7, wherein,

Updating the network resource topology information of the network based on the acquired network resource state information.
The method according to any one of claims 1 to 8, wherein when managing the computing resources and network resources of the network, the method includes:

Perform OAM operations on the computing power resources and network resources of the network;

and / or,

Perform operational processing on the computing resources and/or network resources of the network.
The method according to any one of claims 1 to 8, wherein said managing network computing services includes at least one of the following:

Manage computing power service images;

Manage instances of computing services;

Manage the resources corresponding to the computing power service.
The method according to claim 10, wherein said managing the computing power service image includes at least one of the following:

Added new computing power service image;

Update the version of the computing power service image;

Delete the computing service image.
The method according to claim 10, wherein said managing instances of computing power services includes at least one of the following:

Notifying the second node to establish a connection between nodes and a connection between nodes and terminals;

Evaluate the service quality of the computing power service based at least on the computing power resource status information of the network, the network resource status information and the computing power service status information;

Dispatch the computing power service to at least one fourth node to perform at least one of the following operations:

Instantiate the computing power service;

Update the instance of computing power service;

Termination of computing services;

Wherein, the fourth node has at least a computing power function.
The method according to claim 10, wherein said managing resources corresponding to computing power services includes at least one of the following:

Reserve resources for computing services;

Allocate resources for computing services;

Release resources for computing services;

Expand and/or shrink the resources of the computing power service.
A management and scheduling device, comprising:

The first management unit is configured to manage computing resources and network resources of the network;

The second management unit is configured to manage the computing service of the network;

The scheduling unit is configured to receive the service request of the first service, and schedule the first service.
A node, including: a processor and a communication interface; wherein,

The processor is configured to manage network computing resources and network resources; manage network computing services; and receive a service request for a first service through the communication interface, and schedule the first service .
A node comprising: a processor and memory configured to store a computer program capable of running on the processor,

Wherein, the processor is configured to execute the steps of the method according to any one of claims 1 to 13 when running the computer program.
A storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 13 are realized.