CN119484264A

CN119484264A - A network recommendation method, device, electronic device, chip and medium

Info

Publication number: CN119484264A
Application number: CN202410987861.7A
Authority: CN
Inventors: 赵�怡
Original assignee: China Mobile Communications Group Co Ltd; China Mobile Suzhou Software Technology Co Ltd
Current assignee: China Mobile Communications Group Co Ltd; China Mobile Suzhou Software Technology Co Ltd
Priority date: 2024-07-23
Filing date: 2024-07-23
Publication date: 2025-02-18

Abstract

The disclosure provides a network recommendation method, a network recommendation device, electronic equipment, a chip and a medium, and relates to the technical field of cloud computing. The network recommendation method comprises the steps of collecting network characteristic factors of a bare metal server in multiple dimensions, wherein the multiple dimensions comprise at least two of data traffic, network networking, network configuration, overlay network Overlay and tenant service, and recommending a target position of an access network to the bare metal server through a network recommendation model based on the network characteristic factors. According to the technical scheme provided by the disclosure, the problem that the tenant in the bare metal server cannot obtain the optimal network is solved, and the user experience of the tenant using the bare metal server is improved.

Description

Network recommendation method and device, electronic equipment, chip and medium

Technical Field

The disclosure relates to the field of cloud computing, and in particular relates to a network recommendation method, a device, electronic equipment, a chip and a medium.

Background

The computing resource pool falls the computing instance selection to a physical server node, and the optimal computing node is scheduled according to computing attributes such as a central processor, a memory, a hard disk, affinity and the like of the server. The network or network card of the computing example can be planned in advance only according to the physical connection line, and the user can only select the port of the network equipment accessed by the server, so that the optimal network can not be obtained in the bare metal server.

Disclosure of Invention

The present disclosure aims to solve, at least to some extent, one of the technical problems in the related art.

The disclosure provides a network recommendation method, a device, electronic equipment, a chip and a medium, which are used for solving the problem that a tenant in a bare metal server cannot obtain an optimal network, and automatically recommending optimal network dynamic intelligent recommending and scheduling capability taking user service as a center to a client through correlation analysis of dimensions such as forwarding flow, networking, configuration, overlay network, tenant service and the like, so that the utilization rate of resources and energy sources can be improved, and better user experience is provided for the user.

An embodiment of a first aspect of the present disclosure provides a network recommendation method, including:

collecting network characteristic factors of the bare metal server in multiple dimensions, wherein the multiple dimensions comprise at least two of data traffic, network networking, network configuration, overlay network Overlay and tenant service;

Recommending a target position of the access network to the bare metal server through a network recommendation model based on the network characteristic factors.

In one embodiment of the present disclosure, recommending a target location of an access network to a bare metal server through a network recommendation model based on a network feature factor comprises:

Based on the network feature factors, determining network similarity one by one with each case feature in a case library of a network recommendation model through a similarity algorithm, wherein the case library comprises feature factors of the existing network links of the network where the bare metal server is located;

taking the maximum value in the set formed by the network similarities as the target similarity, and taking the network position corresponding to the target similarity as the target position;

And accessing the bare metal server to the target position.

In one embodiment of the disclosure, before each case feature in the case library of the network recommendation model is determined to be similar to the network one by one through a similarity algorithm based on the network feature factor, the method further comprises the following steps;

Recording different network links in a network where the bare metal server is located;

converting the network link into a network characteristic factor with multiple dimensions;

classifying different network links through a clustering algorithm based on network characteristic factors with multiple dimensions to obtain classification results;

And adding the classification result into the case library.

In one embodiment of the present disclosure, collecting network feature factors of a bare metal server in multiple dimensions includes:

Collecting flow parameters of a network card of a bare metal server and a switch connected with the bare metal server as network characteristic factors of data flow dimension, wherein the flow parameters comprise real-time flow of the network card and the switch, bandwidth of a target port of the network card accessed to the switch and a MAC address table of the switch;

determining the tenant number and network access point of all ports of the switch according to the MAC address table;

Acquiring tenant priorities of all tenants of the bare metal server;

Acquiring service priority and online state of each tenant of all tenants in the tenant service dimension;

Collecting network bandwidth and time delay of tenants in the Overlay network Overlay dimension;

collecting states of all ports of the switch in a network networking dimension;

Collecting the state of a network card in a network configuration dimension;

And taking the flow parameters, the bandwidth of the target port, the tenant number of all ports, the network access point, the tenant priority, the service priority, the online state, the network bandwidth, the time delay, the state of all ports and the state of the network card as network characteristic factors of the bare metal server.

In one embodiment of the present disclosure, collecting network bandwidth and latency of a tenant includes:

and dynamically dial and measure the network bandwidth and the time delay value of the tenant in the Overlay network Overlay dimension through ping and trace.

In one embodiment of the present disclosure, recommending, based on the network feature factor, a target location of an access network to a bare metal server through a network recommendation model, further includes:

Determining a selectable target network list in a virtual local area network or a virtual subnet of the bare metal server based on the service priority and the state of the network card;

Determining the access point position of the target network in the selectable target network list according to the bandwidth of the target port, the real-time flow of the switch and the tenant number of all ports;

the access point location is taken as the target location.

An embodiment of a second aspect of the present disclosure proposes a network recommendation device, including:

The system comprises an acquisition module, a network configuration module and a service management module, wherein the acquisition module is used for acquiring network characteristic factors of the bare metal server in multiple dimensions, wherein the multiple dimensions comprise data traffic, network networking, network configuration, overlay network Overlay and tenant service;

And the recommending module is used for recommending the target position of the access network to the bare metal server through the network recommending model based on the network characteristic factors.

An embodiment of a third aspect of the present disclosure proposes an electronic device comprising at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of the embodiments of the first aspect of the present disclosure.

An embodiment of a fourth aspect of the present disclosure proposes a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method of the embodiment of the first aspect of the present disclosure.

A fifth aspect embodiment of the present disclosure proposes a computer program product comprising a computer program which, when executed by a processor, implements the method of any of the embodiments of the first aspect of the present disclosure.

In summary, according to the network recommendation method provided by the disclosure, network characteristic factors of a bare metal server in multiple dimensions are collected, the multiple dimensions comprise at least two of data traffic, network networking, network configuration, overlay network Overlay and tenant service, data preparation is provided for network recommendation of tenants of the bare metal server, a target position of an access network is recommended to the bare metal server through a network recommendation model based on the network characteristic factors, an optimal network position or path is provided for the tenants in the bare metal server, and network use experience of the tenants is improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure and do not constitute an undue limitation on the disclosure.

FIG. 1 is a schematic diagram of a docking architecture and a service flow of a server in a related art cloud scenario;

FIG. 2 is a schematic diagram of a related art server nanotube TOR switch;

FIG. 3 is a flow chart of a network recommendation method according to an embodiment of the present disclosure;

FIG. 4 is a flowchart of a method for recommending a target location of an access network to a bare metal server via a network recommendation model based on network feature factors in an embodiment of the present disclosure;

Fig. 5 is a flowchart of a case library for constructing a network recommendation model according to an embodiment of the present disclosure;

FIG. 6 is a flow chart of a method of collecting network feature factors for bare metal servers in multiple dimensions in accordance with an embodiment of the present disclosure;

Fig. 7 is a flowchart of a method for collecting network bandwidth and latency of a tenant according to an embodiment of the present disclosure;

FIG. 8 is a flow chart of recommending a target location of an access network to a bare metal server through a network recommendation model based on network characterization factors in an embodiment of the present disclosure;

Fig. 9 is a schematic diagram of network recommendation for a tenant in a bare metal server according to an embodiment of the disclosure;

Fig. 10 is a flowchart of a network recommendation method according to an embodiment of the disclosure;

FIG. 11 is a schematic diagram of a decision tree model according to an embodiment of the present disclosure;

Fig. 12 is a schematic diagram of implementing network traffic load balancing in a bare metal server according to an embodiment of the disclosure;

fig. 13 is an intent to implement tenant number load balancing in a bare metal server according to an embodiment of the disclosure;

Fig. 14 is a schematic structural diagram of a network recommendation device according to an embodiment of the disclosure;

fig. 15 is a block diagram of an electronic device for implementing the network recommendation method of the present disclosure, according to an example embodiment.

Detailed Description

Embodiments of the present disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals identify the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and intended for the purpose of explaining the present disclosure and are not to be construed as limiting the present disclosure.

Related words in this disclosure will first be briefly described:

Bare metal (Bare Metal, BM) server in this disclosure refers to a computing service with both cloud service resilience and physical machine performance. The bare metal server combines the advantages of the traditional physical server and the modern cloud computing, and provides a high-performance, safe and flexible computing resource. The method supports the direct running of codes on hardware, avoids performance loss caused by virtualization, and ensures the monopolization and isolation of resources.

A software defined network (Software Defined Network, SDN), referred to in this disclosure as an emerging network architecture, is one implementation of network virtualization. The core technology OpenFlow separates the control surface from the data surface of the network equipment, so that flexible control of network flow is realized, the network becomes more intelligent, and a good platform is provided for innovation of the core network and application.

A Top of Rack (TOR), in this disclosure, refers to a network device that implements data exchanges in a computer network. Typically at the top of a data center or machine room.

Fig. 1 is a schematic diagram of a docking architecture and a service flow of a server in a related art cloud scenario. In the cloud scene, the SDN is accessed to the bare metal server, and the TOR switch is managed by the SDN controller to realize the switching of the deployment network and the tenant network of the bare metal access as required. Wherein, the butt joint structure includes following main points:

1. The console interfaces with the network pool SDN, and the console interfaces with Nova and Ironic to deploy bare metal. The Nova is one module in openstack software and used for managing virtual machines, ironic is one module in openstack software and used for managing bare metal servers, and openstack is a set of open-source cloud computing management platform.

2. Ironic is responsible for lifecycle management of bare metal servers and deployment of bare metal servers, SDN unifies the nano-tube vSwitch and TOR switches, providing connectivity and switching of deployment and tenant networks of bare metal servers for Ironic.

3. And the tenant applies for the bare metal server according to the specification, and the intercommunication of the virtual machine, the bare machine and the container network under the same virtual private cloud (Virtual Private Cloud, VPC) is realized.

The business process comprises the following steps:

1. In SDN, an administrator network topology discovers device ports and monitors port bandwidth traffic conditions of network devices.

2. An administrator registers the bare metal server, and self-checks and collects the connection relation between the server network card and the switch.

3. Tenant 1 applies for a VPC network, creating a virtual network (network 1) and a virtual subnet (subnet 1).

4. And the tenant applies for the bare metal according to the specification, and selects a network card subnet or vlan of the bare metal server.

5. The tenant network randomly selects 1 network card to create a port (port) of the bare metal server in Nova or Ironic, and updates the port.

6. On SDN, automatically opening a deployment network and a tenant network accessed by a bare metal server, and switching to the tenant network after the bare metal server is automatically installed and successfully deployed. The virtual local area network (Virtual Local Area Network, VLAN) is converted to a virtual extended local area network (VXLAN) by TOR or VXLAN tunnel endpoints (VXLAN Tunnel Endpoint, VTEP), thereby opening up the bare metal server network. The tenant's application may use bare metal servers and virtual machine networks under the same VPC to interwork.

In the related art, a bare metal server is applied and a network card is selected, and the application of a tenant can realize network connection between a Virtual Machine (VM) server and a Bare Metal (BM) server of the same VPC.

Fig. 2 is a schematic diagram of a server nanotube TOR switch in the related art. In the prior art, a computing resource pool drops computing instance selection to a physical server node, and currently, optimal computing nodes are scheduled according to computing attributes such as a central processing unit (Central Processing Unit, CPU), memory, hard disk, affinity and the like of the server. The network or network card of the example can be calculated, the port of the network equipment accessed by the server can be selected by the user only according to the physical connection line planning in advance, and the optimal network scheme can not be recommended and scheduled according to the latest service priority and the service flow carried actually. For example, the server 1 distributes data through the TOR switch 1 of the channel 1, the server 1 distributes data through the TOR switch 2 of the channel 2, and when the channel 1 is busy and the channel 2 is idle, the server can only distribute data according to the fixed channel 1 and the channel 2, and cannot partially shift or equalize the load when the channel 1 is busy to the channel 2 currently in idle.

In the cloud computing field, for life cycle management of a bare metal server, there are optimal scheduling algorithms and schemes in scheduling and selecting computing storage (CPU, storage, hard disk), and in management of a network, related technologies include the following steps to implement network allocation of tenants in the bare metal server:

1. administrator registration switch

1) The switch is connected and is on line.

2) And collecting the port and link information of the switch.

2. Manager registers bare metal server

1) And registering the bare metal server and specification information.

2) And collecting the bare metal server network card, and accessing the link.

3. Tenant applies for bare metal server

1) And applying for the bare metal server by the tenant according to the specification, and deploying the bare metal server.

2) And (5) carrying out automatic installation deployment of bare metal, and switching to a tenant network.

4. Tenant randomly selects a network

1) And selecting 1 subnet1|vlan1 network card at random.

2) The tenant selects the network segment and vlan, and the location where the SDN opens the network is currently the access location of 1 network card randomly selected from the multiple network cards.

5. Tenant use bare metal server

1) The tenant uses a bare metal server.

The above steps can only provide networks for two kinds of tenants of the bare metal server, and for the network to be selected only by the tenants, the SDN cannot provide recommendations of the optimal network in the bare metal server for the tenants to use.

The method is applied to network recommendation tasks in the bare metal server, has rich scenes, and can help tenants to select the best network configuration in cloud service providers and large-scale data centers in cloud computing and data center scenes so as to optimize performance and cost. This is particularly important for resource intensive applications requiring high performance computing, large data analysis, artificial intelligence training, and the like. In a Content Delivery Network (CDN) scenario, a CDN provider may use this technique to automatically select an optimal network path for the content provider to ensure fast, reliable content delivery, improving user experience. In financial services scenarios, such as financial transactions and online paytables, automatically recommending an optimal network can ensure low latency and high reliability of transactions, improving overall performance of the system and user satisfaction. In the gaming industry, online gaming and electronic competitive platforms can provide a smoother gaming experience through automated network optimization, reducing delays and chunking, attracting more players. In the fields of remote education and remote medical treatment, the automatic recommendation optimal network can provide more stable communication quality, and smooth transmission of teaching contents and medical services is ensured. In the internet of things (IoT) field, ioT devices and platforms can improve the stability of connections and the efficiency of data transmission through automatic network optimization, especially in large-scale IoT deployments. For enterprises and government institutions, automatically recommending optimal networks can help them optimize internal network structures, improving office efficiency and data security. In disaster recovery and business continuity planning scenarios, automatic network optimization may ensure that critical applications and services can run over the optimal network path at any time. In the fields of broadcast television and online live broadcast, the automatic recommendation of the optimal network can provide high-quality real-time video transmission, and buffering and interruption are reduced. In the field of electronic commerce, an e-commerce platform can improve the loading speed and transaction processing capacity of a website through automatic network optimization, so that the user experience and sales are improved. In general, these scenarios require fast, reliable and secure network connections, and automatically recommending an optimal network can just meet these needs, improving quality of service and customer satisfaction. The application scenario or application field is not limited in the embodiments of the present disclosure.

The network recommendation method provided by the present disclosure is described in detail below with reference to the accompanying drawings.

Fig. 3 is a flowchart of a network recommendation method according to an embodiment of the present disclosure. The method is performed at an SDN. As shown in the embodiment of fig. 3, the network recommendation method includes:

Step 301, collecting network characteristic factors of a bare metal server in multiple dimensions, wherein the multiple dimensions comprise at least two of data traffic, network networking, network configuration, overlay network Overlay and tenant service.

In this embodiment, the network feature factors refer to indexes used to describe and measure various attributes on the network nodes or edges in the network analysis, and these feature factors can reflect the topology structure, dynamic behavior and functional characteristics of the network. The present disclosure describes a current network by network feature factors of multiple dimensions, including at least two of data traffic, network networking, network configuration, overlay network Overlay, and tenant traffic, where network networking refers to connecting and configuring multiple computers or devices into one network capable of communicating and sharing resources with each other, including local area networks, wide area networks, and wireless networks. Network configuration refers to the specific setup of network devices (e.g., routers, switches, firewalls, etc.) to accommodate a particular network environment and needs. An Overlay network Overlay refers to one or more virtual logical networks built by virtualization technology over an underlying physical network. Tenant traffic refers to various applications and services running in bare metal servers that belong to a particular tenant. To make a network recommendation, it is necessary to collect network feature factors of the bare metal server in multiple dimensions as preparation data.

Step 302, recommending a target position of the access network to the bare metal server through a network recommendation model based on the network characteristic factors.

In this embodiment, the network recommendation model refers to a method for recommending an optimal network location or path to a tenant by using a network feature factor based on a machine learning algorithm, and includes a Case-based reasoning (Case-Based Reasoning, CBR) algorithm. The target location of the access network refers to the optimal network location that the tenant of the bare metal server can currently access. After the network characteristic factors are obtained, the access network and the target position are recommended to the bare metal server through a network recommendation model, and an optimal network position or path can be provided for the tenant.

Fig. 4 is a flowchart of a method for recommending a target location of an access network to a bare metal server through a network recommendation model based on network feature factors according to an embodiment of the present disclosure. Fig. 4 is a further illustration of step 302 of fig. 3, based on the embodiment shown in fig. 4, comprising the steps of:

step 401, based on the network feature factors, determining network similarity with each case feature in a case library of the network recommendation model one by one through a similarity algorithm, wherein the case library comprises feature factors of the existing network links of the network where the bare metal server is located.

In this embodiment, the network recommendation model stores a case library formed by a large number of past case samples, where the case library includes feature factors of existing network links of the network where the bare metal server is located. According to the recorded current network feature factors, network similarity is determined one by one with each case feature in a case library of the network recommendation model through a similarity algorithm, and the obtained network similarity can form a similarity set.

In step 402, the maximum value in the set of network similarities is taken as the target similarity, and the network position corresponding to the target similarity is taken as the target position.

In this embodiment, the maximum value in the set of network similarities is taken as the target similarity, and the network position corresponding to the target similarity is taken as the target position.

And step 403, accessing the bare metal server to the target position.

In this embodiment, the network card of the bare metal server is accessed to the target location. The tenant in the bare metal server can obtain the current optimal network position or path, and balance of network resources in the bare metal server is realized.

Fig. 5 is a flowchart of a case library for constructing a network recommendation model according to an embodiment of the present disclosure. Fig. 5 is a description of the embodiment shown in fig. 5, which is performed before step 401 of fig. 4, and includes the following steps:

Step 501, recording different network links in a network where a bare metal server is located.

In this embodiment, different network links in the network where the bare metal server is located, that is, different network links formed by the network nodes located at the upper layer of the bare metal server in the network topology and each virtual network node in the bare metal server are recorded.

Step 502 converts a network link into a multi-dimensional network feature factor.

In this embodiment, the recorded data of different network links are converted into network feature factors with multiple dimensions, the network link data includes monitoring the switch in a preset time period, collecting port traffic variation and port traffic occupancy rate, analyzing the online positions and specific distribution quantity statistics of the tenant bare metal server and the virtual machine server by collecting a media access Control (MEDIA ACCESS Control, MAC) table of the switch port, collecting the network card state of the bare metal server and the switch port state, and the online state of the associated tenant service, measuring the bandwidth and time delay of the network by a network measuring tool, and counting the tenant service and network segment distribution quantity and the online quantity of the connection of the tenant service. The above collected, measured or counted data is converted into network feature factors of multiple dimensions, namely, the network feature factors are formed in the dimensions of at least two of data traffic, network networking, network configuration, overlay network and tenant service.

Step 503, classifying different network links through a clustering algorithm based on the network characteristic factors of multiple dimensions to obtain classification results.

In this embodiment, the network feature factors of multiple dimensions are clustered, and optionally, feature data of different network links are classified by K nearest neighbors (K Nearest Neighbour, KNN), so as to obtain different classification results.

And step 504, adding the classification result to the case library.

In this embodiment, the classification result obtained by clustering is added to a case library of the network recommendation model, so as to provide a control case for network recommendation in the future. And the lessees can conveniently obtain more refined and accurate network link recommendation.

Fig. 6 is a flowchart of a method for collecting network feature factors of a bare metal server in multiple dimensions according to an embodiment of the disclosure. Fig. 6 is a specific illustration of step 301 of fig. 3, based on the embodiment shown in fig. 6, comprising the steps of:

Step 601, collecting flow parameters of a network card of a bare metal server and a switch connected with the bare metal server, wherein the flow parameters are used as network characteristic factors of data flow dimension, and include real-time flow of the network card and the switch, bandwidth of a target port of the network card accessed to the switch and an MAC address table of the switch.

In this embodiment, the flow parameters of the network card of the bare metal server and the switch connected to the bare metal server are collected, where the flow parameters include real-time flows of the network card and the switch of the bare metal server, a bandwidth of a target port of the network card accessed to the switch, and a MAC address table of the switch. Namely, the real-time flow and port bandwidth indexes of the ports of the switch and the bare metal server network card are obtained.

Step 602, determining the tenant number and network access point of all ports of the switch according to the MAC address table.

In this embodiment, according to the MAC address table, the number of tenants and network access points of all ports of the switch may be obtained, and the number of port users of the network feature factor and the index of the terminal access point are obtained.

Step 603, acquiring tenant priorities of all tenants of the bare metal server.

In this embodiment, the tenant priority refers to a priority allocated to the tenant according to the level of the tenant in the bare metal server. For example, if the tenant is a gold user, or if the tenant rents for a long time and a large amount of money, the tenant has a higher priority. And acquiring tenant priority indexes of all tenants in a user management database of the system.

In step 604, in the tenant service dimension, the service priority and the online status of each tenant of all tenants are obtained.

In this embodiment, in the tenant service dimension, a service priority of each tenant preset by the system and an online state of each tenant are obtained.

Step 605, in the Overlay network Overlay dimension, the network bandwidth and time delay of the tenant are collected.

In this embodiment, network bandwidth and delay values of tenants are collected in the Overlay network Overlay dimension. Network bandwidth and time delay of each tenant in the virtual network are obtained.

In step 606, the states of all ports of the switch are collected in the networking dimension of the network.

In this embodiment, in the network networking dimension, the names of servers connected to the switch and the states of all ports of the switch are collected.

In step 607, the state of the network card is collected in the network configuration dimension.

In this embodiment, in the network configuration dimension, the state of the network card is collected, including the state of the network card device of the bare metal server or the virtual machine server, the state of the network card device of the virtual subnet or the virtual local area network, and the like.

Step 608, the traffic parameter, the bandwidth of the target port, the number of tenants of all ports, the network access point, tenant priority, service priority, on-line status, network bandwidth, time delay, status of all ports and status of the network card are used as network feature factors of the bare metal server.

In this embodiment, the flow parameters collected in the above steps, the bandwidth of the target port, the number of tenants of all ports, the network access point, the tenant priority, the service priority, the online state, the network bandwidth, the time delay, the states of all ports and the states of the network card are used as the network characteristic factors of the bare metal server, so as to provide case characteristic data for the case library of the network recommendation model, and at the same time, provide input data for the network recommendation model, so as to perform optimization recommendation on the network.

Fig. 7 is a flowchart of a method for collecting network bandwidth and time delay of a tenant according to an embodiment of the disclosure. Fig. 7 is a specific illustration of step 605 of fig. 6, based on the embodiment shown in fig. 7, comprising the steps of:

Step 701, dynamically dial and measure network bandwidth and delay values of the tenant in the Overlay network Overlay dimension through ping and trace.

In this embodiment, ping refers to a basic network command for detecting connectivity between two network nodes, checking whether a host is reachable by sending an ICMP echo request and waiting for a response, and measuring the round trip time of a packet. trace refers to tracking and displaying the IP addresses of all routers traversed by a packet during its transmission from a source host to a target host, as well as the delay time for each transmission. And dynamically dial and measure network bandwidth and time delay values of tenants in the Overlay network Overlay dimension through a ping and trace measuring tool. And acquiring network information of the tenant in the bare metal server in the Overlay dimension.

Fig. 8 is a flowchart of recommending a target location of an access network to a bare metal server through a network recommendation model based on network feature factors according to an embodiment of the present disclosure. Fig. 8 is another illustration of step 302 of fig. 3, based on the embodiment shown in fig. 8, comprising the steps of:

Step 801, determining a list of selectable target networks in a virtual local area network or a virtual subnet of the bare metal server based on the service priority and the status of the network card.

In this embodiment, the selectable target network list refers to a list of network links or locations in the vlan or the subnet, where the list includes an optimal network location or path. And according to the service priority of the tenant and the state of the network card in the bare metal server, listing a selectable target network list in the virtual local area network or the virtual sub-network.

Step 802, determining the access point position of the target network in the selectable target network list according to the bandwidth of the target port, the real-time traffic of the switch and the tenant number of all ports.

In this embodiment, according to the bandwidth of the target port, the real-time traffic of the switch, and the number of tenants of all ports of the switch, the optimal access point position of the current target network is determined in the selectable target network list.

Step 803, take the access point location as the target location.

In this embodiment, the location of the access point of the target network is taken as the target location, so as to provide an optimal network for the tenant. The use experience of the tenant on the network in the bare metal server is improved.

In one implementation of this embodiment,

Fig. 9 is a schematic diagram of network recommendation for a tenant in a bare metal server according to an embodiment of the disclosure. As shown in fig. 9, the MAC address of the bare metal server (bare machine) and the port of the switch to which the bare machine is connected are collected, the MAC address of the container in the bare metal server and the host node are collected, and the MAC address of the virtual machine (virtual machine) in the bare metal server and the host node are collected. Based on the network data collected by the bare computers, the containers and the virtual computers, a service is provided by using a network evaluation recommendation model, and optimal network recommendation is provided for tenants. The optimal network recommendation model consists of model factors and a KNN algorithm, wherein the model factors comprise four types, namely service, configuration, networking and flow. The service types include tenant 1, application 1, virtual network1, virtual subnet1, and virtual local area network vlan1. Configuration types include port1 of the bare metal server, virtual network1, virtual local area network vlan1, switch device1, and all ports of the switch device_port1. The networking type comprises an upper host name host connected with the bare metal server, a network card of the bare metal server and a port. The traffic type includes a device name device1 of the switch device or bare metal server device, all ports device_port1 of the switch, a switch port bandwidth (bw) of the switch to bare metal server connection, a number of data packets rate_counter, and a MAC address. The method comprises the steps of carrying out model training on a cloud by using a data set formed by four types of variables through service, configuration, networking and flow weighted summation and a KNN algorithm to obtain a case library of a network recommendation model, carrying out online reasoning at the edge, and intelligently deciding an optimal network scheme. And carrying out decision input according to the data of the network scheme, and carrying out network parameter distribution on the tenant and the application through a Neutron network based on parameters of the tenant, the application, the virtual network, the virtual sub-network and the virtual local area network. And updating ports, virtual networks, virtual local area networks, bare metal servers and switches and ports of the bare metal servers and switches to complete network configuration of tenants. The continuous monitoring of the network is completed through the bare metal server and the switch, the ports of the bare metal server and the switch, the port bandwidth, the number of data packets and the MAC address. The configuration of the optimal network of the tenant using the bare metal server and the updating of the optimal network of other time or other tenants are completed, and therefore the recommendation of the optimal network can be obtained by the tenant.

Fig. 10 is a flowchart of a network recommendation method according to an embodiment of the disclosure. As shown in fig. 10, for an administrator, it includes the steps of:

1. administrator registration switch

1) The switch is connected and is on line.

2) And collecting the port and link information of the switch.

3) The mac table learned by the switch port is monitored.

4) Traffic statistics of switch ports are monitored.

2. Manager registers bare metal server

1) And registering the bare metal server and specification information.

2) And collecting the bare metal server network card, and accessing the link.

3) And collecting the bandwidth and state information of the bare metal server network card.

For the tenant, the method comprises the following steps:

1. tenant applies for bare metal server

2) The bare metal server is deployed by an automatic installation machine and is switched to the tenant network.

2. Tenant randomly selects a network

1) Selecting a tenant network, namely selecting 1 optimal subnet1, vlan1, host1 and bandwidth occupancy network card.

2) The tenant selects the network segment and vlan, and the location of the SDN opening network selects the access location of 1 optimal network card from the plurality of network cards.

3. Tenant use bare metal server

The tenant uses a bare metal server.

The overall architecture of the optimal network recommendation is as follows:

1. Analysis on data plane

1) Find the port of the network card access switch (network card mac- > switch port- > switch uplink port) of the bare metal server.

2) Judging whether the port is idle or not according to the mac table for monitoring the bandwidth, flow statistics and port learning of the port of the switch, and obtaining an idle index.

3) Therefore, the network card 1 of the bare metal server can be found to have a busy switch, and the network card 2 has an idle switch.

2. Control plane analysis

1) The number of tenants and applications that the current network vlan, subnet have used is found.

2) And calculating a weighted index according to the tenant grade and the service grade.

3) And calculating a weighted index according to the configuration use frequency of the network vlan and the subnet.

3. Optimal network recommendation

1) Firstly, according to the service level and configuration of the control plane tenant, the currently available optimal network vlan or subnet is calculated.

2) The location of this network instantiation (switch port or bare metal network card) is then recommended based on the bandwidth, traffic, number of users actually carrying the data plane.

FIG. 11 is a schematic diagram of a decision tree model in accordance with an embodiment of the present disclosure. And recommending optimal positions and paths through a case-based reasoning (CBR+similarity) algorithm (KNN/K-D tree/Bayesian), and selecting paths with small data plane forwarding flow and small state connection and port mac users in a certain time period. And selecting the positions with high tenant service priority, less network segment number use and less service online connection. The network configuration of the selected service is less, the number of service connections is less, and the port is positioned at the position with less online mac number in a certain time period. Through the decision tree model, the network characteristic factors comprise tenant priority, service priority, tenant service configuration, real-time traffic of equipment, port bandwidth, port user quantity, association analysis of terminal access points and equipment configuration service and comprehensive weighting. As shown in fig. 11, the tenant priority of the service is traversed in full quantity to obtain a tenant high priority and a tenant low priority, and policy optimization is performed on the tenant high priority and the tenant low priority in forwarding traffic, configuration state, network delay, service configuration number and online user number, if the forwarding traffic is less than 80%, the forwarding traffic is more likely to be selected (yes), otherwise, the forwarding traffic is more likely to be not selected (no), which means that in a period of time, the unoccupied bandwidth of the data channel is more, and the network recommendation may recommend the network with smaller forwarding traffic to the tenant. For the configuration state, online is preferred over offline ofline. For network latency, if less than 0.1ms, then network recommendations recommend network with low network latency to the tenant. For the number of service configurations, if the number of configured services is less than 10, the network recommendation recommends a network with a small number of service configurations to the tenant in order to obtain more resources to run the application or service of the tenant. For the number of online users, if less than 20, the network recommendation recommends a network with a small number of online users to the tenant, facilitating more network resources available to the tenant. According to comprehensive weighting of forwarding flow, configuration state, network delay, service configuration quantity and online user quantity, making a group decision, judging conflict strategies by using preset values or conditions, and obtaining preference selection of network recommendation, thereby obtaining network recommendation weight based on the decision tree and a preset rule script, storing the network recommendation weight in a case library of a network recommendation model, taking a network access point of a case with highest similarity with the network recommendation weight in network data acquired in real time as an access point of a recommended optimal network, and improving use experience of tenants on a network in a bare metal server.

Fig. 12 is a schematic diagram of implementing network traffic load balancing in a bare metal server according to an embodiment of the disclosure. As shown in fig. 12, for the top-of-rack switches TOR1 and TOR2, the traffic distributed by the SDN server 1 is 100 and 0, and by using the network recommendation method provided by the present disclosure, the network of TOR2 can be recommended to the tenant, so that quick network communication can be performed by using TOR2 without traffic, and the situation that the traffic in TOR1 is fully loaded is reduced, so that load balancing of the network traffic between TOR1 and TOR2 is realized.

Fig. 13 is an intent to implement tenant number load balancing in a bare metal server according to an embodiment of the disclosure. As shown in fig. 13, for the top-of-rack switches TOR1 and TOR2, server 1 distributes 3 tenants 1,2, 3 for TOR1, while TOR2 is not used by the tenant. After the network recommendation is performed on the tenant 3, the tenant 3 accesses the network into the network of the TOR2, so that faster network use experience is obtained, meanwhile, the network load of the TOR1 is lightened, and the network load balance of the TOR1 and the TOR2 is realized.

The embodiment of the disclosure provides a network recommendation method, which is used for collecting network characteristic factors of a bare metal server in multiple dimensions, wherein the multiple dimensions comprise at least two of data traffic, network networking, network configuration, coverage network Overlay and tenant service, providing data preparation for network recommendation for tenants of the bare metal server, recommending a target position of an access network to the bare metal server through a network recommendation model based on the network characteristic factors, providing an optimal network position or path for the tenants in the bare metal server, and improving network use experience of the tenants.

Corresponding to the methods provided in the foregoing several embodiments, the present disclosure further provides a network recommendation device, and since the devices provided in the embodiments of the present disclosure correspond to the methods provided in the foregoing several embodiments, implementation manners of the methods are also applicable to the devices provided in the embodiments, and will not be described in detail in the embodiments.

Fig. 14 is a schematic structural diagram of a network recommendation device 1400 according to an embodiment of the disclosure. As shown in fig. 14, the network recommendation device includes:

The collection module 1410 is configured to collect network feature factors of the bare metal server in multiple dimensions, where the multiple dimensions include data traffic, network networking, network configuration, overlay network Overlay, and tenant service;

a recommendation module 1420 for recommending the target location of the access network to the bare metal server through a network recommendation model based on the network characteristic factor.

In some embodiments, recommendation module 1420 is to:

And accessing the bare metal server to the target position.

In some embodiments, the recommendation module 1420 is further configured to, based on the network feature factors, determine network similarities, one by one, with each case feature in the case base of the network recommendation model via a similarity algorithm;

And adding the classification result into the case library.

In some embodiments, the acquisition module 1410 is to:

Acquiring tenant priorities of all tenants of the bare metal server;

collecting states of all ports of the switch in a network networking dimension;

Collecting the state of a network card in a network configuration dimension;

In some embodiments, the collection module 1410 collects network bandwidth and latency of the tenant as follows:

In some embodiments, the recommendation module 1420 is further to:

the access point location is taken as the target location.

In summary, the network recommendation device is used for collecting network characteristic factors of the bare metal server in multiple dimensions, wherein the multiple dimensions comprise at least two of data traffic, network networking, network configuration, overlay network Overlay and tenant service, and the target position of the access network is recommended to the bare metal server through the network recommendation model based on the network characteristic factors. The device solves the problem that the tenant in the bare metal server cannot obtain the optimal network, and improves the user experience of the tenant using the bare metal server.

In the embodiments provided in the present disclosure, the method and the apparatus provided in the embodiments of the present disclosure are described. In order to implement the functions in the methods provided in the embodiments of the present disclosure, the electronic device may include a hardware structure, a software module, and implement the functions in the form of a hardware structure, a software module, or a hardware structure plus a software module. Some of the functions described above may be implemented in a hardware structure, a software module, or a combination of a hardware structure and a software module.

Fig. 15 is a block diagram of an electronic device 1500 for implementing the network recommendation method described above, according to an example embodiment.

For example, electronic device 1500 may be a mobile phone, computer, messaging device, game console, tablet device, medical device, exercise device, personal digital assistant, or the like.

Referring to FIG. 15, an electronic device 1500 may include one or more of a processing component 1502, a memory 1504, a power component 1506, a multimedia component 1508, an audio component 1510, an input/output (I/O) interface 1512, a sensor component 1514, and a communication component 1516.

The processing component 1502 generally controls overall operation of the electronic device 1500, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 1502 may include one or more processors 1520 to execute instructions to perform all or part of the steps of the methods described above. Further, the processing component 1502 may include one or more modules that facilitate interactions between the processing component 1502 and other components. For example, the processing component 1502 may include a multimedia module to facilitate interaction between the multimedia component 1508 and the processing component 1502.

The memory 1504 is configured to store various types of data to support operations at the electronic device 1500. Examples of such data include instructions for any application or method operating on electronic device 1500, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 1504 may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.

The power supply assembly 1506 provides power to the various components of the electronic device 1500. The power supply component 1506 can include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the electronic device 1500.

The multimedia component 1508 comprises a screen between the electronic device 1500 and the user that provides an output interface. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may sense not only the boundary of a touch or sliding action, but also the duration and pressure associated with the touch or sliding operation. In some embodiments, multimedia assembly 1508 includes a front camera and/or a rear camera. When the electronic device 1500 is in an operational mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

The audio component 1510 is configured to output and/or input audio signals. For example, the audio component 1510 includes a Microphone (MIC) configured to receive external audio signals when the electronic device 1500 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may be further stored in the memory 1504 or transmitted via the communication component 1516. In some embodiments, the audio component 1510 further comprises a speaker for outputting audio signals.

The I/O interface 1512 provides an interface between the processing component 1502 and peripheral interface modules, which can be keyboards, click wheels, buttons, and the like. These buttons may include, but are not limited to, a home button, a volume button, an activate button, and a lock button.

The sensor assembly 1514 includes one or more sensors for providing status assessment of various aspects of the electronic device 1500. For example, the sensor assembly 1514 may detect an on/off state of the electronic device 1500, a relative positioning of the components, such as a display and keypad of the electronic device 1500, the sensor assembly 1514 may also detect a change in position of the electronic device 1500 or a component of the electronic device 1500, the presence or absence of a user in contact with the electronic device 1500, an orientation or acceleration/deceleration of the electronic device 1500, and a change in temperature of the electronic device 1500. The sensor assembly 1514 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 1514 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1514 may also include an acceleration sensor, a gyroscopic sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 1516 is configured to facilitate communication between the electronic device 1500 and other devices, either wired or wireless. The electronic device 1500 may access a wireless network based on a communication standard, such as WiFi,2G or 3G,4G LTE, 5G NR (New Radio), or a combination thereof. In one exemplary embodiment, the communication component 1516 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication component 1516 further includes a Near Field Communication (NFC) module to facilitate short range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, ultra Wideband (UWB) technology, bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the electronic device 1500 may be implemented by one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic elements for executing the methods described above.

In an exemplary embodiment, a non-transitory computer-readable storage medium is also provided, such as memory 1504, including instructions executable by processor 1520 of electronic device 1500 to perform the above-described methods. For example, the non-transitory computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

Embodiments of the present disclosure also provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the network recommendation method described in the above embodiments of the present disclosure.

Embodiments of the present disclosure also provide a computer program product comprising a computer program which, when executed by a processor, performs the network recommendation method described in the above embodiments of the present disclosure.

It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the foregoing figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the disclosure described herein may be capable of operation in sequences other than those illustrated or described herein. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.

In the description of the present specification, reference is made to the description of the terms "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., meaning that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present disclosure. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present disclosure in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present disclosure.

Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, system that includes a processing module, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include an electrical connection (a method of controlling) having one or more wires, a portable computer diskette (a magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium upon which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

It should be understood that portions of embodiments of the present disclosure may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of techniques known in the art, discrete logic circuits with logic gates for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.

Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, and the program may be stored in a computer readable storage medium, where the program when executed includes one or a combination of the steps of the method embodiments.

Furthermore, functional units in various embodiments of the present disclosure may be integrated into one processing module, or each unit may exist alone physically, or two or more units may be integrated into one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented as software functional modules and sold or used as a stand-alone product. The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like.

While embodiments of the present disclosure have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the present disclosure, and that variations, modifications, alternatives, and variations of the above embodiments may be made by those of ordinary skill in the art within the scope of the present disclosure.

Claims

1. A network recommendation method, the method comprising:

recommending a target position of an access network to the bare metal server through a network recommendation model based on the network characteristic factors.

2. The method of claim 1, wherein recommending, based on the network characterization factor, a target location of an access network to the bare metal server via a network recommendation model comprises:

based on the network feature factors, determining network similarity one by one with each case feature in a case library of the network recommendation model through a similarity algorithm, wherein the case library comprises feature factors of the existing network links of the network where the bare metal server is located;

Taking the maximum value in the set formed by the network similarities as a target similarity, and taking the network position corresponding to the target similarity as the target position;

and accessing the bare metal server to the target position.

3. The method of claim 2, further comprising, prior to said determining network similarity, one by one, by a similarity algorithm, with each case feature in the case base of the network recommendation model based on the network feature factor;

converting the network link into a network feature factor of the plurality of dimensions;

Classifying the different network links through a clustering algorithm based on the network characteristic factors with multiple dimensions to obtain classification results;

and adding the classification result to the case library.

4. The method of claim 1, wherein the collecting network characteristic factors of the bare metal server in multiple dimensions comprises:

Collecting flow parameters of a network card of the bare metal server and a switch connected with the bare metal server as the network characteristic factors of the data flow dimension, wherein the flow parameters comprise real-time flow of the network card and the switch, bandwidth of a target port of the network card accessed to the switch and a MAC address table of the switch;

acquiring tenant priorities of all tenants of the bare metal server;

collecting network bandwidth and time delay of the tenant in the Overlay dimension;

Collecting states of all ports of the switch in the network networking dimension;

Collecting the state of the network card in the network configuration dimension;

And taking the flow parameters, the bandwidths of the target ports, the tenant numbers of all ports, the network access point, the tenant priorities, the service priorities, the online status, the network bandwidths, the time delay, the statuses of all ports and the status of the network card as the network characteristic factors of the bare metal server.

5. The method of claim 4, wherein the collecting network bandwidth and latency of the tenant comprises:

and dynamically dial-measuring the network bandwidth and the time delay value of the tenant in the Overlay dimension of the Overlay network through ping and trace.

6. The method of claim 4, wherein recommending, based on the network characterization factor, a target location of an access network to the bare metal server via a network recommendation model, further comprises:

Determining the access point position of a target network in the selectable target network list according to the bandwidth of the target port, the real-time flow of the switch and the tenant number of all ports;

and taking the access point position as the target position.

7. A network recommendation device, the device comprising:

The system comprises an acquisition module, a network configuration module and a service management module, wherein the acquisition module is used for acquiring network characteristic factors of a bare metal server in a plurality of dimensions, wherein the plurality of dimensions comprise data traffic, network networking, network configuration, overlay network Overlay and tenant service;

And the recommending module is used for recommending the target position of the access network to the bare metal server through a network recommending model based on the network characteristic factors.

8. An electronic device, comprising:

At least one processor, and

A memory communicatively coupled to the at least one processor, wherein,

The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

9. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-6.

10. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any of claims 1-6.