Switch load balancing test method and system based on cloud computing

Abstract

The invention discloses a switch load balancing test method and system based on cloud computing, and relates to the technical field of switch load balancing, and the method comprises the steps: calculating a dynamic load evaluation coefficient of each server and a switch server load mean value, and when the switch server load mean value is smaller than 0.8, determining that the switch server load mean value is smaller than 0.8; distributing the new request to the server with the minimum dynamic load evaluation coefficient for processing; when the load mean value of the switch server is not less than 0.8 and less than 1, calculating a hardware processing evaluation coefficient of each server, calculating a second allocation weight of each server, and allocating a new request to the server with the minimum second allocation weight for processing; and when the server load mean value of the switch is not less than 1, calculating a fault processing evaluation coefficient of each server, calculating a third distribution weight of each server, distributing a new request to the server with the minimum third distribution weight for processing, and sending a switch overload early warning to the outside. The load of each server can be more balanced.

Description

Technical Field

[0001] The present invention relates to the technical field of switch load balancing, and in particular to a switch load balancing testing method and system based on cloud computing. Background Art

[0002] A switch is one of the core devices in a computer network. It is responsible for forwarding data packets from one port to another to achieve data exchange and transmission. In a cloud computing environment, switches not only need to have high-speed data forwarding capabilities, but also need to support complex network protocols and intelligent traffic management functions to meet the cloud computing platform's requirements for high performance, high availability, and high scalability. Load balancing is a technology that reasonably allocates network or computing resources to each node to achieve efficient resource utilization and system performance optimization. In a cloud computing environment, load balancing technology ensures that each server node can operate in the optimal state by dynamically adjusting the request allocation strategy, thereby avoiding single point overload and performance bottlenecks.

[0003] In the Chinese invention application with application publication number CN117240787A, a switch load balancing method, device, electronic device and readable storage medium are disclosed, including obtaining user load balancing request information; generating a load balancing configuration flow table according to the user load balancing request information; analyzing and processing the load balancing configuration flow table to obtain a load balancing selection strategy; and performing load balancing scheduling processing on the switch according to the load balancing selection strategy.

[0004] In the above invention application, in the process of load balancing scheduling processing of the switch according to the load balancing selection strategy, when the load balancing selection strategy is a polling load balancing strategy, the preset available port array is polled to determine the first sending port, and the data packet to be sent in the switch is sent through the first sending port; when the load balancing selection strategy is a weighted polling load balancing strategy, the preset available port array is weighted polled to determine the second sending port, and the data packet to be sent in the switch is sent through the second sending port; when the load balancing selection strategy is a hash load balancing strategy, the preset available port array is hashed and determined to determine the third sending port, and the data packet to be sent in the switch is sent through the third sending port. Through the determination of the load balancing selection strategy, different load balancing strategies can be selected, so that the load balancing process can be more flexible. The user only needs to select according to the actual situation, which brings great convenience and flexibility, but the user still needs to set the load balancing strategy independently. The user needs to have certain professional knowledge of network technology and load balancing strategy to correctly set the load balancing strategy. Once the load balancing strategy is set, the user also needs to regularly monitor and adjust the strategy to adapt to changes in the network and application. This requires a lot of time and effort, increasing maintenance costs. At the same time, if the policy is not adjusted properly, it may cause performance degradation or service interruption.

[0005] To this end, the present invention provides a switch load balancing test method and system based on cloud computing. Summary of the invention

[0006] 1. Technical issues to be resolved In view of the shortcomings of the prior art, the present invention provides a switch load balancing test method and system based on cloud computing. The present invention analyzes the switch server load average , automatically selecting different load balancing strategies can make the load of each server more balanced, avoiding the situation where some servers are overloaded while other servers are idle. This helps to improve the stability and response speed of the entire system, thereby solving the technical problems recorded in the background technology.

[0007] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: A switch load balancing test method based on cloud computing includes the following steps: Use a network traffic generator to generate simulated network traffic and periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing; When the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing; When the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

[0008] Furthermore, a network traffic generator is used to generate simulated network traffic, and a network monitoring tool is used to periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server : in, Indicates the serial number of each server connected to the switch. , Indicates the total number of servers connected to the switch.

[0009] Network monitoring tools are software used to monitor and analyze network traffic, performance, and security. They provide network administrators with a window to observe and understand network activities and ensure the stability and security of the network. They include Wireshark: an open source network protocol analyzer that can capture and view network packets in real time. It supports multiple protocol analysis and has powerful filtering and search capabilities. Nagios: an open source network monitoring system that can monitor network services, host resources, etc. It has custom plug-ins and extensions and supports event processing and notification functions. PRTG: a comprehensive network monitoring tool that provides a variety of sensors for monitoring network traffic, bandwidth, etc. It has an instant alarm and notification system and supports remote monitoring via mobile applications. Zabbix: An enterprise-class open source network monitoring solution that supports monitoring of various network devices, servers, and virtual machines. Distributed monitoring architecture, advanced alarm and notification functions, and support for a variety of charts and reports. OpManager: A comprehensive network monitoring and management tool that can monitor network performance, servers, and virtual environments. Real-time network monitoring and troubleshooting capabilities, integrated alarm and event management capabilities. SolarWinds NPM: A powerful network monitoring and management tool that can monitor network performance in real time. Provides detailed network topology and alarm functions, and supports historical performance data storage and report generation.

[0010] Furthermore, the dynamic weight load evaluation coefficient of each server is obtained , calculate the average load of the switch server : When the switch server load average When it is less than 0.8, the switch distributes requests according to the first balancing strategy, and distributes new requests to the server with the smallest dynamic load evaluation coefficient for processing.

[0011] Furthermore, when the switch server load average If the value is not less than 0.8 and less than 1, check the server information to obtain the number of processor cores of each server connected to the switch. , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server : in, Indicates the maximum number of processor cores of all servers connected to the switch. Indicates the maximum value of the main frequency of all servers connected to the switch. Indicates the maximum memory capacity of all servers connected to the switch.

[0012] The number of processor cores refers to the number of physical computing units in the processor that independently perform tasks. Each core can independently process data and complete computing tasks. The number of cores directly affects the number of tasks that the processor can handle simultaneously. Multi-core processors can handle multiple tasks at the same time, improving overall processing efficiency.

[0013] The processor frequency is an indicator that describes the processor's operating speed. It indicates the speed at which the digital pulse signal inside the processor oscillates. The higher the frequency, the more instructions the processor can execute per unit time, thereby increasing the processing speed.

[0014] Furthermore, the hardware processing evaluation coefficient of each server is obtained and dynamic load rating factor , calculate the second allocation weight of each server : in, Indicates the minimum value of the hardware processing evaluation coefficient of all servers connected to the switch. Indicates the maximum value of the dynamic load evaluation coefficients of all servers connected to the switch.

[0015] Furthermore, when the switch server load average When it is not less than 1, search the fault log to obtain the historical fault frequency of all servers connected to the switch. , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server : in, Indicates the maximum historical failure frequency of all servers connected to the switch. Indicates the maximum value of the recovery time objective of all servers connected to the switch. Indicates the maximum value of the recovery point objective of all servers connected to the switch.

[0016] RTO is the maximum tolerable amount of time a computer, system, network, or application will be out of service following a failure or disaster. It represents the time it takes to recover from an outage (whether planned, unplanned, or caused by a disaster) and resume normal operations for an application or set of applications.

[0017] RPO refers to the age of files that must be restored from backup storage to ensure normal operation of the system when a computer, system or network fails due to hardware, program or communication failure. It defines the maximum amount of data loss that an enterprise can tolerate when restoring data from the most recent valid backup after a catastrophic event such as data loss or system crash, that is, the tolerable data loss time window.

[0018] Further, obtain the second allocation weight of each server and fault handling evaluation coefficient , calculate the third allocation weight of each server : in, Indicates the minimum value of the fault handling evaluation coefficient of all servers connected to the switch.

[0019] The cloud computing-based switch load balancing test system includes: The first balanced distribution module uses a network traffic generator to generate simulated network traffic and periodically detects the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing; The second balancing module, when the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing; The third balanced distribution module, when the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

[0020] (III) Beneficial effects The present invention provides a switch load balancing test method and system based on cloud computing, which has the following beneficial effects: 1. Use a network traffic generator to generate simulated network traffic and periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch will assign new requests to the server with the smallest dynamic load evaluation coefficient. Allocating new requests to the server with the smallest dynamic load evaluation coefficient can make the load of each server more balanced and avoid the situation where some servers are overloaded while other servers are idle. This helps to improve the stability and response speed of the entire system.

[0021] 2. When the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimum server processing helps ensure that when the server cluster is under high load, new requests can be assigned to servers with relatively strong processing power, thereby making full use of hardware resources and making the load of each server more balanced. This helps to avoid some servers from crashing due to overload while ensuring that other servers can continue to operate stably.

[0022] 3. When the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective (RTO) and Recovery Point Objective (RPO) , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The smallest server processes and sends out switch overload warnings, which helps ensure that under high load conditions, servers with relatively strong processing power and better fault recovery capabilities can take on more requests, thereby improving the overall stability and response speed of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of the flow of a switch load balancing test method based on cloud computing of the present invention; Figure 2 It is a structural schematic diagram of a switch load balancing test system based on cloud computing of the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] See also Figure 1 The present invention provides a switch load balancing test method based on cloud computing, comprising the following steps: Step 1: Use a network traffic generator to generate simulated network traffic and periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing.

[0026] The step 1 includes the following contents: Step 101: Use a network traffic generator to generate simulated network traffic, and use a network monitoring tool to periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server : in, Indicates the serial number of each server connected to the switch. , Indicates the total number of servers connected to the switch.

[0027] Network monitoring tools are software used to monitor and analyze network traffic, performance, and security. They provide network administrators with a window to observe and understand network activities and ensure the stability and security of the network. They include Wireshark: an open source network protocol analyzer that can capture and view network packets in real time. It supports multiple protocol analysis and has powerful filtering and search capabilities. Nagios: an open source network monitoring system that can monitor network services, host resources, etc. It has custom plug-ins and extensions and supports event processing and notification functions. PRTG: a comprehensive network monitoring tool that provides a variety of sensors for monitoring network traffic, bandwidth, etc. It has an instant alarm and notification system and supports remote monitoring via mobile applications. Zabbix: An enterprise-class open source network monitoring solution that supports monitoring of various network devices, servers, and virtual machines. Distributed monitoring architecture, advanced alarm and notification functions, and support for a variety of charts and reports. OpManager: A comprehensive network monitoring and management tool that can monitor network performance, servers, and virtual environments. Real-time network monitoring and troubleshooting capabilities, integrated alarm and event management capabilities. SolarWinds NPM: A powerful network monitoring and management tool that can monitor network performance in real time. Provides detailed network topology and alarm functions, and supports historical performance data storage and report generation.

[0028] Step 102: Obtain the dynamic weight load evaluation coefficient of each server , calculate the average load of the switch server : When the switch server load average When it is less than 0.8, the switch distributes requests according to the first balancing strategy, and distributes new requests to the server with the smallest dynamic load evaluation coefficient for processing.

[0029] When using, combine the contents in steps 101 and 102: Use a network traffic generator to generate simulated network traffic and periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch will assign new requests to the server with the smallest dynamic load evaluation coefficient. Allocating new requests to the server with the smallest dynamic load evaluation coefficient can make the load of each server more balanced and avoid the situation where some servers are overloaded while other servers are idle. This helps to improve the stability and response speed of the entire system.

[0030] Step 2: When the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing.

[0031] The step 2 includes the following contents: Step 201: When the switch server load average If the value is not less than 0.8 and less than 1, check the server information to obtain the number of processor cores of each server connected to the switch. , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server : in, Indicates the maximum number of processor cores of all servers connected to the switch. Indicates the maximum value of the main frequency of all servers connected to the switch. Indicates the maximum memory capacity of all servers connected to the switch.

[0032] The number of processor cores refers to the number of physical computing units in the processor that independently perform tasks. Each core can independently process data and complete computing tasks. The number of cores directly affects the number of tasks that the processor can handle simultaneously. Multi-core processors can handle multiple tasks at the same time, improving overall processing efficiency.

[0033] The processor frequency is an indicator that describes the processor's operating speed. It indicates the speed at which the digital pulse signal inside the processor oscillates. The higher the frequency, the more instructions the processor can execute per unit time, thereby increasing the processing speed.

[0034] Step 202: Obtain the hardware processing evaluation coefficient of each server and dynamic load rating factor , calculate the second allocation weight of each server : in, Indicates the minimum value of the hardware processing evaluation coefficient of all servers connected to the switch. Indicates the maximum value of the dynamic load evaluation coefficients of all servers connected to the switch.

[0035] When the switch server load average When it is not less than 0.8 and less than 1, the switch distributes requests according to the second balancing strategy, and distributes new requests to the second distribution weight. Minimal server processing.

[0036] When using, combine the contents in steps 201 and 202: When the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimum server processing helps ensure that when the server cluster is under high load, new requests can be assigned to servers with relatively strong processing power, thereby making full use of hardware resources and making the load of each server more balanced. This helps to avoid some servers from crashing due to overload while ensuring that other servers can continue to operate stably.

[0037] Step 3: When the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

[0038] The step three includes the following contents: Step 301: When the switch server load average When it is not less than 1, search the fault log to obtain the historical fault frequency of all servers connected to the switch. , Recovery Time Objective (RTO) and Recovery Point Objective (RPO) , calculate the fault handling evaluation coefficient of each server : in, Indicates the maximum historical failure frequency of all servers connected to the switch. Indicates the maximum value of the recovery time objective (RTO) of all servers connected to the switch. Indicates the maximum value of the recovery point objective (RPO) of all servers connected to the switch.

[0039] RTO is the maximum tolerable amount of time a computer, system, network, or application will be out of service following a failure or disaster. It represents the time it takes to recover from an outage (whether planned, unplanned, or caused by a disaster) and resume normal operations for an application or set of applications.

[0040] RPO refers to the age of files that must be restored from backup storage to ensure normal operation of the system when a computer, system or network fails due to hardware, program or communication failure. It defines the maximum amount of data loss that an enterprise can tolerate when restoring data from the most recent valid backup after a catastrophic event such as data loss or system crash, that is, the tolerable data loss time window.

[0041] Step 302: Obtain the second allocation weight of each server and fault handling evaluation coefficient , calculate the third allocation weight of each server : in, Indicates the minimum value of the fault handling evaluation coefficient of all servers connected to the switch.

[0042] When the switch server load average When it is not less than 1, the switch distributes requests according to the third balancing strategy, and distributes new requests to the third distribution weight. The server processes the smallest amount of data and sends out switch overload warnings.

[0043] When used, combine the contents in steps 301 and 302: When the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective (RTO) and Recovery Point Objective (RPO) , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The smallest server processes and sends out switch overload warnings, which helps ensure that under high load conditions, servers with relatively strong processing power and better fault recovery capabilities can take on more requests, thereby improving the overall stability and response speed of the system.

[0044] See also Figure 2 The present invention provides a switch load balancing test system based on cloud computing, comprising: The first balanced distribution module uses a network traffic generator to generate simulated network traffic and periodically detects the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing; The second balancing module, when the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing; The third balanced distribution module, when the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

[0045] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0046] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0047] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A switch load balancing test method based on cloud computing, characterized in that: The steps include: Use a network traffic generator to generate simulated network traffic and periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing; When the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing; When the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

2. The switch load balancing test method based on cloud computing according to claim 1 is characterized in that: Use a network traffic generator to generate simulated network traffic, and use a network monitoring tool to periodically detect the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server : in, Indicates the sequence number of each server connected to the switch. , Indicates the total number of servers connected to the switch.

3. The switch load balancing test method based on cloud computing according to claim 2 is characterized in that: Get the dynamic weight load evaluation coefficient of each server , calculate the average load of the switch server : When the switch server load average When it is less than 0.8, the switch distributes requests according to the first balancing strategy, and distributes new requests to the server with the smallest dynamic load evaluation coefficient for processing.

4. The switch load balancing test method based on cloud computing according to claim 3 is characterized in that: When the switch server load average If the value is not less than 0.8 and less than 1, check the server information to obtain the number of processor cores of each server connected to the switch. , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server : in, Indicates the maximum number of processor cores of all servers connected to the switch. Indicates the maximum value of the main frequency of all servers connected to the switch. Indicates the maximum memory capacity of all servers connected to the switch.

5. The switch load balancing test method based on cloud computing according to claim 4 is characterized in that: Get the hardware processing evaluation coefficient of each server and dynamic load rating factor , calculate the second allocation weight of each server : in, Indicates the minimum value of the hardware processing evaluation coefficient of all servers connected to the switch. Indicates the maximum value of the dynamic load evaluation coefficients of all servers connected to the switch.

6. The switch load balancing test method based on cloud computing according to claim 1, characterized in that: When the switch server load average When it is not less than 1, search the fault log to obtain the historical fault frequency of all servers connected to the switch. , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server : in, Indicates the maximum historical failure frequency of all servers connected to the switch. Indicates the maximum value of the recovery time objective of all servers connected to the switch. Indicates the maximum value of the recovery point objective of all servers connected to the switch.

7. The switch load balancing test method based on cloud computing according to claim 6 is characterized in that: Get the second allocation weight for each server and fault handling evaluation coefficient , calculate the third allocation weight of each server : in, Indicates the minimum value of the fault handling evaluation coefficient of all servers connected to the switch.

8. A switch load balancing test system based on cloud computing, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first balanced distribution module uses a network traffic generator to generate simulated network traffic and periodically detects the CPU usage of each server connected to the switch. , memory usage and network bandwidth usage , calculate the dynamic load evaluation coefficient of each server and switch server load average , when the switch server load average When it is less than 0.8, the switch assigns new requests to the server with the smallest dynamic load evaluation coefficient for processing; The second balancing module, when the switch server load average When it is not less than 0.8 and less than 1, obtain the number of processor cores of each server connected to the switch , main frequency and memory capacity , calculate the hardware processing evaluation coefficient of each server , combined with the dynamic load evaluation factor , calculate the second allocation weight of each server , the switch assigns the new request to the second allocation weight Minimal server processing; The third balanced distribution module, when the switch server load average When it is not less than 1, obtain the historical failure frequency of all servers connected to the switch , Recovery Time Objective and recovery point objective , calculate the fault handling evaluation coefficient of each server , combined with the second distribution weight of each server , calculate the third allocation weight of each server , the switch assigns the new request to the third allocation weight The server processes the smallest amount of data and sends out switch overload warnings.

Images