Digital-twin-based distributed system performance dynamic simulation test method and platform

Abstract

The application discloses a distributed system performance dynamic simulation test method and platform based on digital twinning, relates to the technical field of digital twinning, and solves the problem of one-sidedness of distributed system performance test, and comprises the following steps: acquiring each physical node in the distributed system, identifying active nodes in the physical nodes, and collecting node static data of the active nodes; setting corresponding virtual nodes according to the node static data of the active nodes, and constructing a virtual twin system of the distributed system based on the virtual nodes and digital twinning technology; collecting node dynamic information of the active nodes in a monitoring period, and constructing a simulation test environment of the virtual twin system according to the node dynamic information; and performing performance test on the virtual twin system based on the simulation test environment, taking the performance test result of the virtual twin system as the performance test result of the distributed system. The application realizes comprehensive test on the performance of the distributed system by performing performance test on the virtual twin system of the distributed system.

Description

Technical Field

[0001] This invention belongs to the field of digital twin technology, specifically a method and platform for dynamic simulation testing of distributed system performance based on digital twins. Background Technology

[0002] Data twins collect real-time data of physical entities through technologies such as sensors and the Internet of Things, and combine this with 3D modeling, simulation analysis, and other methods to build a dynamically updated digital model in virtual space. This model can not only reflect the current state of the physical entity, but also predict future behavior through algorithms, and even simulate responses in different scenarios. A distributed system is a system in which a group of independent computer nodes are interconnected through a network to work together to complete a specific task. These nodes are physically dispersed, but logically form a unified whole, and achieve resource sharing, load balancing, and fault tolerance through message passing.

[0003] However, when performing performance testing on distributed systems, due to limited resources in the real environment, it is difficult to run multiple test scenarios. At the same time, it is also difficult to actively trigger extreme anomalies in the real environment, which leads to a lag in problem discovery and makes the performance testing of distributed systems rather one-sided.

[0004] To this end, this invention proposes a method and platform for dynamic simulation testing of distributed system performance based on digital twins. Summary of the Invention

[0005] The purpose of this invention is to propose a dynamic simulation testing method and platform for the performance of distributed systems based on digital twins, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A dynamic simulation and testing method for the performance of distributed systems based on digital twins, the method includes:

[0008] Step S1: Obtain each physical node in the distributed system, identify the active nodes among the physical nodes, and collect the static node data of the active nodes;

[0009] Step S2: Based on the static data of active nodes, set up corresponding virtual nodes, and construct a virtual twin system of the distributed system based on the virtual nodes and digital twin technology.

[0010] Step S3: Collect the node dynamic information of active nodes during the monitoring period, and construct the simulation test environment of the virtual twin system based on the node dynamic information;

[0011] Step S4: Perform performance testing on the virtual twin system based on the simulation test environment, and output the performance test results of the virtual twin system as the performance test results of the distributed system.

[0012] Further, step S1 includes the following sub-steps:

[0013] Step S11: Read all physical nodes in the distributed system and record the node IP address corresponding to each physical node;

[0014] Step S12: Identify the host IP address of the host corresponding to the distributed system, randomly generate an application sequence number, and package the application sequence number, host IP address, and node IP address of any physical node to obtain a node access data packet;

[0015] Step S13: Construct a node access request through the host, and send the node access request carrying the node access data packet to the corresponding physical node.

[0016] Step S14: The physical node receives the node access request sent by the host, parses the node access data packet and reads the node IP address in the node access data packet. If the node IP address in the node access data packet is different from the node IP address corresponding to the physical node, no operation is performed.

[0017] If the node IP address in the node access data packet is the same as the node IP address corresponding to the physical node, then the host IP address and the request sequence number are read, the request sequence number is processed to obtain the response sequence number, the node IP address, the host IP address, and the response sequence number are packaged to obtain the node feedback response packet, and the physical node generates a feedback response signal and sends the node feedback response packet back to the host.

[0018] Furthermore, step S1 also includes the following sub-steps:

[0019] Step S15: The host receives the feedback response signal, parses the node feedback response packet, and obtains the response sequence number and host IP address in the node feedback response packet;

[0020] Step S16: Reverse process the response sequence number to obtain the parsed sequence number corresponding to the response sequence number; compare the parsed sequence number with the application sequence number. If the parsed sequence number is different from the application sequence number, no operation is performed.

[0021] If the parsed sequence number is the same as the requested sequence number, the host IP address in the node's response packet is compared with the host IP address corresponding to the host. If the host IP address in the node's response packet is different from the host IP address corresponding to the host, no operation is performed. If the host IP address in the node's response packet is the same as the host IP address corresponding to the host, the corresponding physical node is recorded as an active node.

[0022] Step S17: Count the node IP addresses of all active nodes, and read the operating system version, CPU information, memory information, disk information and network bandwidth of the corresponding active node based on the node IP address, and summarize them into the node static data of the corresponding active node.

[0023] Further, step S2 includes the following sub-steps:

[0024] Step S21: Obtain the static data of active nodes, including the node IP address, operating system version, CPU information, memory information, disk information, and network bandwidth of any active node.

[0025] Step S22: Create a corresponding virtual node for each active node, use the node IP address of the active node as the simulation IP address of the virtual node, and use the operating system version, CPU information, memory information, disk information and network bandwidth as attributes of the virtual node.

[0026] Step S23: Use a 3D visualization modeling tool to construct a 3D visualization model corresponding to the distributed system, set the virtual nodes according to the physical topology of the distributed system, add corresponding interactive interfaces, and obtain a virtual twin system corresponding to the distributed system.

[0027] Further, step S3 includes the following sub-steps:

[0028] Step S31: Collect the CPU utilization, memory usage, network uplink bandwidth, network downlink bandwidth, disk read speed, disk write speed, and interface response time of active nodes in real time through monitoring tools during multiple monitoring periods; the multiple monitoring periods include normal load periods, abnormal load periods, and peak load periods.

[0029] Step S32: Summarize the CPU utilization of the same active node within the monitoring period in chronological order to obtain the time-series CPU utilization of the corresponding active node.

[0030] The memory usage, network uplink bandwidth, network downlink bandwidth, disk read rate, disk write rate, and interface response time of the same active node within the monitoring period are all summarized in chronological order to obtain the time-series memory usage, time-series network uplink bandwidth, time-series network downlink bandwidth, time-series disk read rate, time-series disk write rate, and time-series interface response time of the corresponding active node.

[0031] Furthermore, step S3 also includes the following sub-steps:

[0032] Step S33: Record the timing CPU utilization, timing memory usage, timing network uplink bandwidth, timing network downlink bandwidth, timing disk read rate, timing disk write rate, and timing interface response time of the same active node as the node dynamic information of the corresponding active node; thereby obtaining the node dynamic information of all active nodes.

[0033] Step S34: Inject the node dynamic information of each active node during all normal load cycles into the corresponding virtual node of the virtual twin system to obtain the normal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during all abnormal load cycles into the corresponding virtual node of the virtual twin system to obtain the abnormal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during peak load cycles into the corresponding virtual node of the virtual twin system to obtain the peak load simulation test environment of the virtual twin system.

[0034] Further, step S4 includes the following sub-steps:

[0035] Step S41: Obtain multiple simulation test environments and adjust the virtual twin system to the normal load simulation test environment;

[0036] Step S42: Simulate k concurrent users in a normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation.

[0037] Step S43: Calculate the login response time for each login operation, the browsing response time for each browsing operation, and the data interaction time for each data interaction operation.

[0038] The login response time, browsing response time, and data interaction time are compared with their respective thresholds, and the number of login timeouts, browsing timeouts, and data interaction timeouts are recorded.

[0039] Step S44: Calculate the login timeout rate, browsing timeout rate, and data interaction timeout rate sequentially based on the number of login timeouts, browsing timeouts, and data interaction timeouts; record the login timeout rate, browsing timeout rate, and data interaction timeout rate as the test results of the normal load simulation test environment.

[0040] Furthermore, step S4 also includes the following sub-steps:

[0041] Step S45: Adjust the virtual twin system to the abnormal load simulation test environment. Simulate k concurrent users in the normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation.

[0042] Step S46: Count the number of login operations, browsing operations, and data interaction operations with error status codes; then, divide the number of login operations, browsing operations, and data interaction operations with the corresponding total number of login operations, browsing operations, and data interaction operations to obtain the login operation error rate, browsing operation error rate, and data interaction operation error rate.

[0043] Furthermore, step S4 also includes the following sub-steps:

[0044] Step S47: Adjust the virtual twin system to the peak load simulation test environment, simulate j concurrent users in the peak load simulation test environment, and calculate the test results corresponding to the peak load simulation test environment based on the operations described in steps S42-S44.

[0045] Step S48: Output the test results of the normal load simulation test environment, abnormal load simulation test environment and peak load simulation test environment in the virtual twin system as the performance test results of the distributed system.

[0046] On the other hand, the distributed system performance dynamic simulation test platform based on digital twins executes the distributed system performance dynamic simulation test method based on digital twins. The platform includes a data acquisition module, a node analysis module, a node generation module, a digital twin module, an environment simulation module, and a performance test module.

[0047] The data acquisition module is used to collect the node IP addresses of all physical nodes in the distributed system and send them to the node analysis module; the node analysis module is used to analyze the node activity status of physical nodes, obtain the node IP addresses of all active nodes and feed them back to the data acquisition module; the data acquisition module collects the node static data of active nodes based on the node IP addresses of active nodes and sends it to the node generation module.

[0048] The node generation module is used to set up corresponding virtual nodes based on the static node data of active nodes and send the virtual nodes to the digital twin module; the digital twin module is used to construct a virtual twin system of the distributed system based on the virtual nodes and digital twin technology, and the digital twin module sends the virtual twin system of the distributed system to the environment simulation module; the data acquisition module is also used to collect the node dynamic information of active nodes during the monitoring period and send it to the environment simulation module.

[0049] The environment simulation module is used to construct a simulation test environment for the virtual twin system based on the node dynamic information of active nodes and send it to the performance test module; the performance test module is used to perform performance tests on the virtual twin system based on the simulation test environment, and output the performance test results of the virtual twin system as the performance test results of the distributed system.

[0050] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0051] 1. This invention obtains each physical node in a distributed system, identifies active nodes among the physical nodes, and collects the static node data of the active nodes; sets corresponding virtual nodes based on the static node data of the active nodes, and constructs a virtual twin system of the distributed system based on the virtual nodes and digital twin technology, thereby realizing the construction of the virtual twin system corresponding to the distributed system.

[0052] 2. This invention collects the node dynamic information of active nodes within the monitoring period, constructs a simulation test environment for the virtual twin system based on the node dynamic information, performs performance testing on the virtual twin system based on the simulation test environment, and outputs the performance test results of the virtual twin system as the performance test results of the distributed system, thereby realizing a comprehensive test of the performance of the distributed system. Attached Figure Description

[0053] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0054] Figure 1 This is a flowchart of the method of the present invention;

[0055] Figure 2 This is a flowchart illustrating the identification process of active nodes in this invention.

[0056] Figure 3 This is a schematic diagram of the dynamic simulation test platform in this invention. Detailed Implementation

[0057] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Example 1: Please refer to Figure 1 and Figure 2 As shown, the technical solution provided by this invention is as follows: a dynamic simulation test method for the performance of a distributed system based on digital twins. This method involves collecting active nodes corresponding to the distributed system, mapping virtual nodes based on the static data of the active nodes, and constructing a virtual twin system corresponding to the distributed system using digital twin technology combined with the virtual nodes. A simulation test environment for the virtual twin system is obtained by collecting the dynamic information of each active node over a period of time and importing it into the virtual nodes. Performance simulation tests are then performed on the virtual twin system within the simulation test environment, and the performance simulation test results of the virtual twin system are used as the performance test results of the distributed system and output.

[0059] In this invention, the distributed system performance dynamic simulation test method is specifically as follows:

[0060] Step S1: Obtain each physical node in the distributed system, identify the active nodes among the physical nodes, and collect the static node data of the active nodes;

[0061] In this context, a physical node is an independent, real hardware computing unit in a distributed system; specifically, a physical node is a physical entity that carries system software and data, usually referring to a physical server; unless otherwise specified in this invention, a physical node is assumed to be a physical server.

[0062] In this invention, step S1 includes the following sub-steps:

[0063] Step S11: Read all physical nodes in the distributed system and record the node IP address corresponding to each physical node;

[0064] Step S12: Identify the host IP address of the host corresponding to the distributed system, randomly generate an application sequence number, and package the application sequence number, host IP address, and node IP address of any physical node to obtain a node access data packet;

[0065] The application sequence number is a string with randomly generated length and content; the node IP address depends on the physical node to which the node access data packet is expected to be sent; specifically, the node access data packet is a data packet.

[0066] Step S13, as follows Figure 2As shown, a node access request is constructed by the host, and the host sends the node access request carrying the node access data packet to the corresponding physical node.

[0067] Step S14: The physical node receives the node access request sent by the host, parses the node access data packet and reads the node IP address in the node access data packet. If the node IP address in the node access data packet is different from the node IP address corresponding to the physical node, no operation is performed.

[0068] If the node IP address in the node access data packet is the same as the node IP address corresponding to the physical node, then the host IP address and the request sequence number are read, the request sequence number is processed to obtain the response sequence number, the node IP address, the host IP address, and the response sequence number are packaged to obtain the node feedback response packet, and the physical node generates a feedback response signal and sends the node feedback response packet back to the host.

[0069] Specifically, the processing of the application serial number involves performing a pre-defined string operation on the application serial number. Taking the simplest operation as an example, a "1" is added to the end of the application serial number, and the application serial number after the addition is recorded as the response serial number.

[0070] Step S15: The host receives the feedback response signal, parses the node feedback response packet, and obtains the response sequence number and host IP address in the node feedback response packet;

[0071] Step S16: Perform the processing steps of step S14 in reverse order on the response sequence number to obtain the parsed sequence number corresponding to the response sequence number; compare the parsed sequence number with the application sequence number. If the parsed sequence number is different from the application sequence number, no operation is performed.

[0072] If the parsed sequence number is the same as the requested sequence number, the host IP address in the node's response packet is compared with the host IP address corresponding to the host. If the host IP address in the node's response packet is different from the host IP address corresponding to the host, no operation is performed. If the host IP address in the node's response packet is the same as the host IP address corresponding to the host, the corresponding physical node is recorded as an active node.

[0073] Step S17: Count the node IP addresses of all active nodes, and read the operating system version (operating system type and version number), CPU information (number of cores and clock frequency), memory information (total available memory), disk information (disk capacity and type), and network bandwidth (node ​​bandwidth limit) of the corresponding active node based on the node IP address, and summarize them into the node static data of the corresponding active node.

[0074] Step S2: Based on the static data of active nodes, set up corresponding virtual nodes, and construct a virtual twin system of the distributed system based on the virtual nodes and digital twin technology.

[0075] In this invention, step S2 includes the following sub-steps:

[0076] Step S21: Obtain the static data of active nodes, including the node IP address, operating system version, CPU information, memory information, disk information, and network bandwidth of any active node.

[0077] Step S22: Create a corresponding virtual node for each active node, use the node IP address of the active node as the simulation IP address of the virtual node, and use the operating system version, CPU information, memory information, disk information and network bandwidth as attributes of the virtual node.

[0078] Virtual nodes are abstract mappings of physical nodes. There are multiple ways to create virtual nodes in practice. For example, virtual nodes can be created through containers (such as using Docker to write the runtime environment of virtual nodes and build an image to obtain virtual nodes), virtual nodes can be created through virtual machines (such as creating virtual nodes directly through VirtualBox), and virtual nodes can be created through dedicated simulation tools (such as creating virtual nodes directly through Mininet). In practice, the choice depends on the type of distributed system.

[0079] Step S23: Use a 3D visualization modeling tool to construct a 3D visualization model corresponding to the distributed system, set the layout of the virtual nodes according to the physical topology of the distributed system and add corresponding interactive interfaces to obtain a virtual twin system corresponding to the distributed system.

[0080] The physical topology of the distributed system describes the physical connection method and layout structure of each active node in the system; the construction of the corresponding 3D visualization model of the distributed system can be implemented using tools (such as Unity or Unreal Engine) or low-code platforms (such as MATLAB or Simulink).

[0081] Step S3: Collect the node dynamic information of active nodes during the monitoring period, and construct the simulation test environment of the virtual twin system based on the node dynamic information;

[0082] In this invention, step S3 includes the following sub-steps:

[0083] Step S31: Collect the CPU utilization, memory usage, network uplink bandwidth, network downlink bandwidth, disk read speed, disk write speed, and interface response time of active nodes in real time through monitoring tools during multiple monitoring periods; the multiple monitoring periods include normal load periods, abnormal load periods, and peak load periods.

[0084] Specifically, normal load refers to the state of a distributed system when handling expected, average-level business traffic; abnormal load refers to a distributed system deviating from its normal state due to internal failures or external interference, entering an unhealthy or functionally impaired state; peak load refers to a distributed system experiencing traffic pressure far exceeding normal levels, but the traffic itself is legitimate and expected business traffic.

[0085] Step S32: Summarize the CPU utilization of the same active node within the monitoring period in chronological order to obtain the time-series CPU utilization of the corresponding active node.

[0086] The memory usage, network uplink bandwidth, network downlink bandwidth, disk read rate, disk write rate, and interface response time of the same active node within the monitoring period are all summarized in chronological order to obtain the time-series memory usage, time-series network uplink bandwidth, time-series network downlink bandwidth, time-series disk read rate, time-series disk write rate, and time-series interface response time of the corresponding active node.

[0087] Step S33: Record the timing CPU utilization, timing memory usage, timing network uplink bandwidth, timing network downlink bandwidth, timing disk read rate, timing disk write rate, and timing interface response time of the same active node as the node dynamic information of the corresponding active node; thereby obtaining the node dynamic information of all active nodes.

[0088] Step S34: Inject the node dynamic information of each active node during all normal load cycles into the corresponding virtual node of the virtual twin system to obtain the normal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during all abnormal load cycles into the corresponding virtual node of the virtual twin system to obtain the abnormal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during peak load cycles into the corresponding virtual node of the virtual twin system to obtain the peak load simulation test environment of the virtual twin system.

[0089] Step S4: Perform performance testing on the virtual twin system based on the simulation test environment, and output the performance test results of the virtual twin system as the performance test results of the distributed system.

[0090] In this invention, step S4 includes the following sub-steps:

[0091] Step S41: Obtain multiple simulation test environments and adjust the virtual twin system to the normal load simulation test environment;

[0092] Step S42: Simulate k concurrent users in a normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation. Here, k is a constant, and the distributed system does not reach the peak load period state after adding k concurrent users under normal load period conditions.

[0093] It should be noted that the specific number of times the login, browsing, and data interaction operations mentioned above are executed can be adjusted according to the type and scale of the distributed system; data operations depend on the type of distributed system; for example, for an e-commerce system, the data interaction operation is placing an order.

[0094] Step S43: Calculate the login response time for each login operation, the browsing response time for each browsing operation, and the data interaction time for each data interaction operation; compare the login response time, browsing response time, and data interaction time with the corresponding thresholds, and record the number of login timeouts, browsing timeouts, and data interaction timeouts.

[0095] Step S44: Calculate the login timeout rate, browsing timeout rate, and data interaction timeout rate sequentially based on the number of login timeouts, browsing timeouts, and data interaction timeouts. Record the login timeout rate, browsing timeout rate, and data interaction timeout rate as the test results of the normal load simulation test environment.

[0096] Step S45: Adjust the virtual twin system to the abnormal load simulation test environment. Simulate k concurrent users in the normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation.

[0097] Step S46: Count the number of login operations, browsing operations, and data interaction operations with error status codes; then, divide the number of login operations, browsing operations, and data interaction operations by the total number of login operations, browsing operations, and data interaction operations respectively to obtain the login operation error rate, browsing operation error rate, and data interaction operation error rate.

[0098] The status code is a 3-digit numeric code representing the result of the server's processing of the request. Generally speaking, error status codes include 400, 401, 404, 500, and 502, etc.

[0099] Step S47: Adjust the virtual twin system to the peak load simulation test environment. Simulate j concurrent users in the peak load simulation test environment. Based on the operations described in steps S42-S44, calculate the test results corresponding to the peak load simulation test environment; where j is a constant and j is much greater than k.

[0100] Step S48: Output the test results of the normal load simulation test environment, abnormal load simulation test environment and peak load simulation test environment in the virtual twin system as the performance test results of the distributed system, and provide them as a reference for system design, operation and maintenance and optimization.

[0101] It should be noted that the projects tested in the above simulation test environment are only for example, and adjustments should be made according to the scale and type of the distributed system in actual practice.

[0102] Example 2: Please refer to Figure 3 As shown, based on another concept of the same invention, a dynamic simulation test platform for the performance of a distributed system based on digital twins is proposed, including a data acquisition module, a node analysis module, a node generation module, a digital twin module, an environment simulation module, and a performance test module.

[0103] The data acquisition module is used to collect the node IP addresses of all physical nodes in the distributed system and send them to the node analysis module; the node analysis module is used to analyze the node activity status of physical nodes, obtain the node IP addresses of all active nodes and feed them back to the data acquisition module; the data acquisition module collects the node static data of active nodes based on the node IP addresses of active nodes and sends it to the node generation module.

[0104] The node generation module is used to set up corresponding virtual nodes based on the static node data of active nodes and send the virtual nodes to the digital twin module; the digital twin module is used to construct a virtual twin system of the distributed system based on the virtual nodes and digital twin technology, and the digital twin module sends the virtual twin system of the distributed system to the environment simulation module; the data acquisition module is also used to collect the node dynamic information of active nodes during the monitoring period and send it to the environment simulation module.

[0105] The environment simulation module is used to construct a simulation test environment for the virtual twin system based on the node dynamic information of active nodes and send it to the performance test module; the performance test module is used to perform performance tests on the virtual twin system based on the simulation test environment, and output the performance test results of the virtual twin system as the performance test results of the distributed system.

[0106] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A dynamic simulation and testing method for the performance of distributed systems based on digital twins, characterized in that, The methods include: Step S1: Obtain each physical node in the distributed system, identify the active nodes among the physical nodes, and collect the static node data of the active nodes; Step S2: Based on the static data of active nodes, set up corresponding virtual nodes, and construct a virtual twin system of the distributed system based on the virtual nodes and digital twin technology. Step S3: Collect the node dynamic information of active nodes during the monitoring period, and construct the simulation test environment of the virtual twin system based on the node dynamic information; Step S4: Perform performance testing on the virtual twin system based on the simulation test environment, and output the performance test results of the virtual twin system as the performance test results of the distributed system.

2. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 1, characterized in that, Step S1 includes the following sub-steps: Step S11: Read all physical nodes in the distributed system and record the node IP address corresponding to each physical node; Step S12: Identify the host IP address of the host corresponding to the distributed system, randomly generate an application sequence number, and package the application sequence number, host IP address, and node IP address of any physical node to obtain a node access data packet; Step S13: Construct a node access request through the host, and send the node access request carrying the node access data packet to the corresponding physical node. Step S14: The physical node receives the node access request sent by the host, parses the node access data packet and reads the node IP address in the node access data packet. If the node IP address in the node access data packet is different from the node IP address corresponding to the physical node, no operation is performed. If the node IP address in the node access data packet is the same as the node IP address corresponding to the physical node, then the host IP address and the request sequence number are read, the request sequence number is processed to obtain the response sequence number, the node IP address, the host IP address, and the response sequence number are packaged to obtain the node feedback response packet, and the physical node generates a feedback response signal and sends the node feedback response packet back to the host.

3. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 2, characterized in that, Step S1 further includes the following sub-steps: Step S15: The host receives the feedback response signal, parses the node feedback response packet, and obtains the response sequence number and host IP address in the node feedback response packet; Step S16: Reverse process the response sequence number to obtain the parsed sequence number corresponding to the response sequence number; compare the parsed sequence number with the application sequence number. If the parsed sequence number is different from the application sequence number, no operation is performed. If the parsed sequence number is the same as the requested sequence number, the host IP address in the node's response packet is compared with the host IP address corresponding to the host. If the host IP address in the node's response packet is different from the host IP address corresponding to the host, no operation is performed. If the host IP address in the node's response packet is the same as the host IP address corresponding to the host, the corresponding physical node is recorded as an active node. Step S17: Count the node IP addresses of all active nodes, and read the operating system version, CPU information, memory information, disk information and network bandwidth of the corresponding active node based on the node IP address, and summarize them into the node static data of the corresponding active node.

4. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 1, characterized in that, Step S2 includes the following sub-steps: Step S21: Obtain the static data of active nodes, including the node IP address, operating system version, CPU information, memory information, disk information, and network bandwidth of any active node. Step S22: Create a corresponding virtual node for each active node, use the node IP address of the active node as the simulation IP address of the virtual node, and use the operating system version, CPU information, memory information, disk information and network bandwidth as attributes of the virtual node. Step S23: Use a 3D visualization modeling tool to construct a 3D visualization model corresponding to the distributed system, set the virtual nodes according to the physical topology of the distributed system, add corresponding interactive interfaces, and obtain a virtual twin system corresponding to the distributed system.

5. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 1, characterized in that, Step S3 includes the following sub-steps: Step S31: Collect the CPU utilization, memory usage, network uplink bandwidth, network downlink bandwidth, disk read speed, disk write speed, and interface response time of active nodes in real time through monitoring tools during multiple monitoring periods; the multiple monitoring periods include normal load periods, abnormal load periods, and peak load periods. Step S32: Summarize the CPU utilization of the same active node within the monitoring period in chronological order to obtain the time-series CPU utilization of the corresponding active node. The memory usage, network uplink bandwidth, network downlink bandwidth, disk read rate, disk write rate, and interface response time of the same active node within the monitoring period are all summarized in chronological order to obtain the time-series memory usage, time-series network uplink bandwidth, time-series network downlink bandwidth, time-series disk read rate, time-series disk write rate, and time-series interface response time of the corresponding active node.

6. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 5, characterized in that, Step S3 further includes the following sub-steps: Step S33: Record the timing CPU utilization, timing memory usage, timing network uplink bandwidth, timing network downlink bandwidth, timing disk read rate, timing disk write rate, and timing interface response time of the same active node as the node dynamic information of the corresponding active node; thereby obtaining the node dynamic information of all active nodes. Step S34: Inject the node dynamic information of each active node during all normal load cycles into the corresponding virtual node of the virtual twin system to obtain the normal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during all abnormal load cycles into the corresponding virtual node of the virtual twin system to obtain the abnormal load simulation test environment of the virtual twin system; inject the node dynamic information of each active node during peak load cycles into the corresponding virtual node of the virtual twin system to obtain the peak load simulation test environment of the virtual twin system.

7. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 6, characterized in that, Step S4 includes the following sub-steps: Step S41: Obtain multiple simulation test environments and adjust the virtual twin system to the normal load simulation test environment; Step S42: Simulate k concurrent users in a normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation. Step S43: Calculate the login response time for each login operation, the browsing response time for each browsing operation, and the data interaction time for each data interaction operation. The login response time, browsing response time, and data interaction time are compared with their respective thresholds, and the number of login timeouts, browsing timeouts, and data interaction timeouts are recorded. Step S44: Calculate the login timeout rate, browsing timeout rate, and data interaction timeout rate sequentially based on the number of login timeouts, browsing timeouts, and data interaction timeouts; record the login timeout rate, browsing timeout rate, and data interaction timeout rate as the test results of the normal load simulation test environment.

8. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 7, characterized in that, Step S4 further includes the following sub-steps: Step S45: Adjust the virtual twin system to the abnormal load simulation test environment. Simulate k concurrent users in the normal load simulation test environment. Each concurrent user performs one login operation, ten browsing operations, and one data interaction operation. Step S46: Count the number of login operations, browsing operations, and data interaction operations with error status codes; The error rates for login, browsing, and data interaction operations are obtained by dividing the number of errors in login, browsing, and data interaction operations by the total number of login, browsing, and data interaction operations, respectively.

9. The method for dynamic simulation testing of distributed system performance based on digital twins according to claim 8, characterized in that, Step S4 further includes the following sub-steps: Step S47: Adjust the virtual twin system to the peak load simulation test environment, simulate j concurrent users in the peak load simulation test environment, and calculate the test results corresponding to the peak load simulation test environment based on the operations described in steps S42-S44. Step S48: Output the test results of the normal load simulation test environment, abnormal load simulation test environment and peak load simulation test environment in the virtual twin system as the performance test results of the distributed system.

10. A dynamic simulation and testing platform for the performance of a distributed system based on digital twins, characterized in that, The platform for implementing the distributed system performance dynamic simulation test method based on digital twins according to any one of claims 1-9 includes a data acquisition module, a node analysis module, a node generation module, a digital twin module, an environment simulation module, and a performance test module. The data acquisition module is used to collect the node IP addresses of all physical nodes in the distributed system and send them to the node analysis module; the node analysis module is used to analyze the node activity status of physical nodes, obtain the node IP addresses of all active nodes, and feed them back to the data acquisition module. The data acquisition module collects static node data of active nodes based on their IP addresses and sends it to the node generation module. The node generation module is used to set up corresponding virtual nodes based on the static node data of active nodes and send the virtual nodes to the digital twin module; The digital twin module is used to construct a virtual twin system of the distributed system based on virtual nodes and digital twin technology. The digital twin module sends the virtual twin system of the distributed system to the environment simulation module. The data acquisition module is also used to collect the node dynamic information of active nodes during the monitoring period and send it to the environment simulation module. The environment simulation module is used to construct a simulation test environment for the virtual twin system based on the node dynamic information of active nodes and send it to the performance test module. The performance testing module is used to perform performance testing on the virtual twin system based on a simulation testing environment, and outputs the performance test results of the virtual twin system as the performance test results of the distributed system.

Images