Cloud computer gateway cluster dynamic hot expansion and contraction method and system

Abstract

The application discloses a cloud computer gateway cluster dynamic hot expansion and contraction method and system, belongs to the cloud computing technical field, and aims to solve the technical problem of how to realize dynamic hot expansion and contraction of the gateway cluster scale without affecting the use of current cloud computer users. The method comprises the following steps: configuring index thresholds of cloud computer gateway cluster expansion and contraction through a management platform; integrating a gateway management service and a monitoring service in a cloud computer gateway image; if the current gateway cluster load pressure exceeds the upper limit of the thresholds, expansion operation is performed; if the current gateway cluster load pressure is lower than the lower limit of the dynamic hot expansion and contraction thresholds, contraction operation is performed; according to the load pressure of each gateway node in the gateway cluster, the gateway management service selects the optimal gateway through a load balancing method and returns gateway information to a client, and the client initiates a connection request to the corresponding gateway based on the gateway information to obtain cloud computer transmission protocol flow data.

Description

Technical Field

[0001] This invention relates to the field of cloud computing technology, specifically to a method and system for dynamic hot expansion and contraction of cloud computer gateway clusters. Background Technology

[0002] With the maturity of cloud computing technology, a variety of SaaS applications have evolved at the upper layer. As a new form of computer, cloud computers have solved the pain points of traditional computers, such as data security risks, large workload of operation and maintenance deployment, resource solidification, poor flexibility, and high hardware replacement costs. They have the advantages of flexible office, rapid deployment, unified operation and maintenance management, and data not being stored locally. They have been favored by major public cloud vendors, who have launched their own cloud computer products for users to choose from and sell on the public Internet.

[0003] The traffic fluctuations of cloud PCs are relatively regular. During the day, customer access volume is high with small fluctuations, putting significant pressure on the cloud PC gateway cluster. At night, customer access volume is low with small fluctuations, resulting in less traffic pressure on the cloud PC gateway cluster. Meanwhile, with the launch of various promotional activities, the peak traffic brought by customers during promotional periods is significantly higher than usual. To adapt to the changing traffic fluctuations, improve user experience, reduce the risks caused by connection session fluctuations, and save hardware resource costs, the need for dynamic hot-scaling and scaling of cloud PC gateway clusters has emerged. This is primarily used to address the increased pressure caused by network traffic fluctuations, which leads to a decrease in network transmission quality and product experience for existing users.

[0004] How to achieve dynamic hot scaling of the gateway cluster without affecting the current cloud PC users' usage is a technical problem that needs to be solved. Summary of the Invention

[0005] The technical objective of this invention is to address the above-mentioned shortcomings by providing a method and system for dynamic hot expansion and contraction of cloud computer gateway clusters, thereby solving the technical problem of how to achieve dynamic hot expansion and contraction of the gateway cluster size without affecting the current use of cloud computers by users.

[0006] This invention discloses a method for dynamic hot expansion and contraction of a cloud computer gateway cluster, comprising the following steps:

[0007] Data configuration: Configure the threshold values ​​for scaling up and down the cloud PC gateway cluster through the management platform, including the upper limit, lower limit and step size of the threshold, and specify the cloud PC gateway image used for scaling up and down the cloud PC gateway cluster;

[0008] Service Configuration: The gateway management service and monitoring service are integrated into the cloud PC gateway image. The monitoring service is used to monitor the health status and performance data of the current gateway and report to the cloud PC monitoring platform. The gateway management service is used to select the optimal gateway and establish a connection with the client based on the current gateway cluster load pressure and through load balancing methods.

[0009] Cluster scaling up and down: Based on the current gateway's health status and performance data, cluster scaling up and down is determined. If the current gateway cluster load exceeds the upper threshold, the following scaling up operation is performed: Based on the step size, the management platform uses the cloud PC gateway image to create a new gateway and registers the gateway to the gateway management service. The scaling up operation is repeated until the gateway cluster load is lower than the upper threshold. If the current gateway cluster load is lower than the lower threshold of dynamic hot scaling up and down, the following scaling down operation is performed: The management platform deletes the gateway with the lowest load, and after all existing connections related to the gateway with the lowest load are naturally disconnected, the physical resources of the gateway with the lowest load are deleted. The scaling down operation is repeated until the gateway cluster load is higher than the lower threshold.

[0010] Connection establishment: The client initiates a connection request to the cloud PC gateway management service. Based on the load pressure of each gateway node in the gateway cluster, the gateway management service selects the optimal gateway through a load balancing method and returns the gateway information to the client. The client then initiates a connection request to the corresponding gateway based on the gateway information to obtain the cloud PC transmission protocol stream data.

[0011] As a preferred option, the threshold parameters include gateway bandwidth, gateway CPU load, user reconnection rate, user disconnection rate, user latency, and number of user connections.

[0012] Performance data includes gateway CPU, memory, network stress, frequency of abnormal session closures, and number of session connections.

[0013] As a preferred option, the cloud PC gateway image supports virtualization type images and container type images, including Qcow2 format, Raw format and Docker Image format.

[0014] As a preferred option, when performing cluster scaling up or down, if both scaling up and scaling down operations are triggered simultaneously, the scaling up operation has a higher priority than the scaling down operation. Only the scaling up operation is executed, and the administrator is prompted to modify the configuration of the optimization metric threshold.

[0015] As a preferred option, after the cloud PC gateway service allocates the gateway address, it also sends the gateway address to the client and the internal protocol terminal of the cloud PC. The client and the internal protocol terminal of the cloud PC simultaneously initiate connection requests to the gateway.

[0016] As preferred options, load balancing algorithms include round-robin scheduling, weighted round-robin scheduling, random balancing, weighted random balancing, least-connection scheduling, weighted least-connection scheduling, destination address hashing scheduling, and source address hashing scheduling.

[0017] As a preferred approach, scaling up involves creating virtual machines or containers on new physical nodes to handle gateway connection loads. This is different from and has an anti-affinity relationship with the nodes distributed by existing virtual machines or containers.

[0018] In a second aspect, the present invention provides a dynamic hot-expansion and shrinkage system for a cloud computer gateway cluster, comprising a cloud computer gateway cluster, a management platform, and a client. The cloud computer gateway cluster, the management platform, and the client work together to realize dynamic expansion and shrinkage operations of the cloud computer gateway cluster through a dynamic hot-expansion and shrinkage method for a cloud computer gateway cluster as described in any of the first aspects.

[0019] The cloud PC gateway cluster dynamic hot expansion and contraction method and system of the present invention have the following advantages: It realizes the dynamic hot expansion and contraction of the cloud PC gateway cluster, and can dynamically adjust the cluster size according to indicators such as business pressure and user access without affecting user use. It solves the problems of user lag and disconnection caused by excessive gateway pressure due to changes in business access volume, or resource waste caused by insufficient pressure. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] The invention will be further described below with reference to the accompanying drawings.

[0022] Figure 1 This is a flowchart of a dynamic hot-expansion / shrinkage method for a cloud computer gateway cluster, as described in Example 1. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments are not intended to limit the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0024] This invention provides a dynamic hot-scaling system for cloud PC gateway clusters, which solves the technical problem of how to achieve dynamic hot-scaling of the gateway cluster without affecting the current use of cloud PCs by users.

[0025] Example 1:

[0026] This invention provides a method for dynamic hot scaling of a cloud computer gateway cluster, comprising four steps: data configuration, service configuration, cluster scaling, and connection establishment.

[0027] Step S100 Data Configuration: Configure the threshold values ​​for scaling up and down the cloud PC gateway cluster through the management platform, including the upper limit, lower limit and step size of the threshold, and specify the cloud PC gateway image used for scaling up and down the cloud PC gateway cluster.

[0028] In this embodiment, the administrator configures the index thresholds for scaling up and down the cloud PC gateway cluster on the management platform. The index thresholds include gateway bandwidth, gateway CPU pressure, user reconnection rate, user disconnection rate, user latency, and number of user connections.

[0029] Among them, the cloud PC gateway image supports virtualization type images and container type images, including Qcow2 format, Raw format and Docker Image format.

[0030] Step S200 Service Configuration: Integrate gateway management service and monitoring service within the cloud PC gateway image. The monitoring service is used to monitor the current gateway's health status and performance data and report them to the cloud PC monitoring platform. The gateway management service is used to select the optimal gateway and establish a connection with the client based on the current gateway cluster load pressure and through load balancing methods.

[0031] In this embodiment, performance data includes gateway CPU, memory, network pressure, frequency of abnormal session closure, number of session connections, etc.

[0032] Step S300 Cluster Scaling / Scaling: Based on the current gateway's health status and performance data, cluster scaling / scaling is determined. If the current gateway cluster load exceeds the upper threshold, the following scaling / scaling operation is performed: Based on the step size, the management platform uses the cloud computer gateway image to create a new gateway and registers the gateway to the gateway management service. The scaling / scaling operation is repeated until the gateway cluster load is lower than the upper threshold. If the current gateway cluster load is lower than the lower threshold for dynamic hot scaling / scaling, the following scaling / scaling operation is performed: The management platform deletes the gateway with the lowest load, and after all existing connections related to the gateway with the lowest load are naturally disconnected, the physical resources of the gateway with the lowest load are deleted. The scaling / scaling operation is repeated until the gateway cluster load is higher than the lower threshold.

[0033] Load balancing algorithms include round-robin scheduling, weighted round-robin scheduling, random balancing, weighted random balancing, least-connection scheduling, weighted least-connection scheduling, destination address hash scheduling, and source address hash scheduling.

[0034] In this embodiment, the expansion and contraction of the gateway does not affect the existing cloud computer session connections. After expansion, the newly added gateway will share the existing connection pressure. After contraction, the deleted gateway will no longer add new session connections until all sessions are disconnected and then the resource deletion operation is performed.

[0035] In this embodiment, the scaling operation carries the gateway connection pressure by creating virtual machines or containers on new physical nodes, which are different from and have anti-affinity with the nodes distributed by existing virtual machines or containers.

[0036] In this embodiment, when performing cluster scaling up or down, if both scaling up and scaling down operations are triggered simultaneously, the scaling up operation has a higher priority than the scaling down operation. Only the scaling up operation is executed, and the administrator is prompted to modify the configuration of the optimization indicator threshold.

[0037] Step S400 Connection Establishment: The client initiates a connection request to the cloud computer gateway management service. Based on the load pressure of each gateway node in the gateway cluster, the gateway management service selects the optimal gateway through a load balancing method and returns the gateway information to the client. The client then initiates a connection request to the corresponding gateway based on the gateway information to obtain the cloud computer transmission protocol stream data.

[0038] In this embodiment, after each disconnection, the client needs to re-request the gateway management service, which then allocates a gateway in real time based on the current status.

[0039] After the cloud PC gateway service assigns a gateway address, it also sends the gateway address to the client and the internal protocol terminal of the cloud PC. The client and the internal protocol terminal of the cloud PC simultaneously initiate connection requests to the gateway.

[0040] The method in this embodiment can dynamically adjust the size of the gateway cluster based on the current access volume of cloud PC services and gateway traffic pressure, without affecting the current user experience. This balances the gateway pressure, improves the user experience, saves computing resources, and reduces underlying overhead. It resolves the issues of user lag and disconnections caused by excessive gateway pressure due to changes in service access volume, or resource waste caused by insufficient pressure.

[0041] Example 2:

[0042] The present invention discloses a dynamic hot expansion and contraction system for a cloud computer gateway cluster, comprising a cloud computer gateway cluster, a management platform, and a client. The cloud computer gateway cluster, the management platform, and the client work together to realize dynamic expansion and contraction operations of the cloud computer gateway cluster through the method disclosed in Embodiment 1.

[0043] The present invention has been shown and described in detail above with reference to the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above embodiments, those skilled in the art will know that more embodiments of the present invention can be obtained by combining the means in the different embodiments described above, and these embodiments are also within the protection scope of the present invention.

Claims

1. A cloud computer gateway cluster dynamic hot scaling method, characterized in that, Comprising the following steps: Data configuration: configure the index threshold of cloud computer gateway cluster scaling through the management platform, including threshold upper limit, threshold lower limit and step length, and specify the cloud computer gateway image used by the cloud computer gateway cluster scaling; Service configuration: integrate gateway management service and monitoring service in the cloud computer gateway image, the monitoring service is used to monitor the health status and performance data of the current gateway and report to the cloud computer monitoring platform, and the gateway management service is used to select the optimal gateway through load balancing method based on the current gateway cluster load pressure to establish connection with the client; Cluster scaling: based on the health status and performance data of the current gateway, if the current gateway cluster load pressure exceeds the threshold upper limit, execute the following expansion operation: based on the step length, the management platform creates a new gateway using the cloud computer gateway image and registers the gateway to the gateway management service, and repeats the expansion operation until the gateway cluster load pressure is lower than the threshold upper limit; If the current gateway cluster load pressure is lower than the threshold lower limit of dynamic hot scaling, execute the following scaling operation: the management platform deletes the gateway with the smallest load pressure, after the existing connections related to the gateway with the smallest load pressure are naturally disconnected, delete the physical resources of the gateway with the smallest load pressure, and repeat the scaling operation until the gateway cluster load pressure is higher than the threshold lower limit; Connection establishment: the client initiates a connection request to the cloud computer gateway management service, and the gateway management service selects the optimal gateway through load balancing method based on the load pressure of each gateway node in the gateway cluster, and returns the gateway information to the client, and the client initiates a connection request to the corresponding gateway based on the gateway information to obtain the cloud computer transmission protocol stream data.

2. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, The index threshold includes gateway bandwidth, gateway CPU pressure, user reconnection rate, user disconnection rate, user delay and user connection number; The performance data includes gateway CPU, memory, network pressure, session abnormal closing frequency and session connection number.

3. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, The types of cloud computer gateway image support virtualization type image and container type image, including Qcow2 format, Raw format and DockerImage format.

4. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, When performing cluster scaling, if the expansion operation and the scaling operation are triggered at the same time, the priority of the expansion operation is higher than that of the scaling operation, only the expansion operation is executed, and the administrator is prompted to modify the configuration of the optimization index threshold.

5. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, After the cloud computer gateway service allocates the gateway address, the gateway address is simultaneously sent to the client and the cloud computer internal protocol end, and the client and the cloud computer internal protocol end simultaneously initiate a connection request to the gateway.

6. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, The load balancing algorithm includes round robin scheduling algorithm, weighted round robin scheduling algorithm, random balancing algorithm, weighted random balancing algorithm, minimum connection scheduling algorithm, weighted minimum connection scheduling algorithm, target address hash scheduling algorithm and source address hash scheduling algorithm.

7. The cloud computer gateway cluster dynamic hot scaling method of claim 1, wherein, The expansion operation creates a virtual machine or a container on a new physical node to bear the gateway connection pressure, which is different from the existing virtual machine or container distributed node and is anti-affinity.

8. A cloud computer gateway cluster dynamic hot scaling system, characterized in that, The cloud computer gateway cluster, the management platform and the client are matched to realize the dynamic expansion and contraction operation of the cloud computer gateway cluster by the cloud computer gateway cluster dynamic hot expansion and contraction method in any one of claims 1-7.

Images