CN108038013A - Distributed performance test method and device and computer-readable recording medium - Google Patents

Abstract

The invention discloses a distributed performance testing method and device and a computer-readable storage medium. The distributed performance testing method includes: analyzing and obtaining configuration information of each slave server according to prestored node allocation information, and the configuration information includes network address and the number of nodes allocated by each slave server; according to the configuration information of each slave server, split the total test data to obtain the sub-test data corresponding to each node, and replace the data reference in the test script with the corresponding sub-test data of each node, set up the test script corresponding to each node; push the test program, sub-test data and test script corresponding to each node to the nodes of each slave server; according to the test task, send corresponding commands to the nodes of each slave server , each node executes the task and feeds back the execution result. By implementing the present invention, a test task can be automatically divided and carried out synchronously on multiple nodes, so as to meet the requirements of large concurrent test scenarios.

Description

Distributed performance testing method and device, and computer-readable storage medium

technical field

The invention relates to the testing field, in particular to a distributed performance testing method and device and a computer-readable storage medium.

Background technique

Large concurrent test tasks (over 60,000 virtual users) are often encountered in performance testing, and a single stress source node cannot support such a high test scenario; commercial performance testing tools are relatively expensive, and open source testing tools are used to implement complex business There are certain deficiencies in the scenario, which requires manual modification of test data, test scripts, etc., which is laborious and error-prone.

The test tool Jmeter has a remote running function. Through this module, multiple remote engines can be controlled from a single Jmeter client to simulate a large load on the server. In theory one instance of Jmeter client can control any number of remote Jmeter instances and collect all data from them. The remote operation function is realized by the following means:

Save the test script on the local machine;

Manage multiple Jmeter engines from a single machine;

The client is responsible for sending the content (request) of the script to all servers;

All nodes run the same test plan, and each node runs a complete test plan; moreover, the remote running mode will take up more resources, instead of running the same number of non-GUI tests independently, if too many node instances are used, It will cause Jmeter client resource overload. In addition, for parameterized test scenarios, since each node runs the same test plan, it is necessary to prepare test data files under the same path on each server. Under this limitation, a server can only run one remote instance, which greatly increases the requirements of the test environment.

In the remote operation mode of Jmeter, before executing the test, you need to start the Jmeter instance of each node in Jmeter server mode (manual execution); for the test plan that references the test data, you need to manually split the data file, and then place it in each remote Under the same path of the server, and in this scenario, only one remote instance can run on each server; if there are too many nodes, the resource overhead and network overhead of the Jmeter client will be greatly increased, and the test results may be affected by certain degree of pollution.

During the performance test task of grabbing red envelopes during the Spring Festival, due to the simulation of high concurrency (more than 50,000) scenarios, a single node cannot meet the test requirements, and the business requires that each account can only be called once and cannot be reused. The method adopted at that time was to manually split the test data into multiple parts, place them in different paths of different servers, then copy multiple copies of Jmeter, and rename them respectively and place them in the corresponding paths. Finally, manually modify the test script to make multiple similar copies (the reference content of the test data is different), and then manually switch to different servers, start the Jmeter program separately, and run their respective test scripts independently. The whole process is manually involved. If the test scenario changes, multiple test scripts need to be changed, and if the test data changes, it must be manually re-split and copied to different nodes of different servers. A test scenario that lasts 20 minutes often requires about half an hour of preparation work. The time cost is very high, and each script must be completed manually, which is particularly error-prone. There are six or seven nodes. In case one of them makes an error, the test The task is about to fail and needs to be started from scratch, requiring very high accuracy of human editing.

Contents of the invention

In view of this, the present invention aims to provide a distributed performance testing method, device and computer-readable storage medium, which avoids the problem of overloading the system overhead of the test program client in the scenario of multiple remote nodes, and reduces the contamination of test results risks of.

Specifically, a distributed performance testing method of the present invention includes: analyzing and obtaining configuration information of each slave server according to pre-stored node allocation information, the configuration information including the network address and the number of nodes allocated by each slave server; Describe the configuration information of each slave server, split the total test data, obtain the sub-test data corresponding to each node, replace the data reference in the test script with the sub-test data corresponding to each node, and establish the test script corresponding to each node Push the test program, sub-test data and test scripts corresponding to each node to the nodes of each slave server; according to the test task, send corresponding commands to the nodes of each slave server, and each node executes the task and feeds back the execution result.

Further, the step of splitting the total test data to obtain the sub-test data corresponding to each node also includes: configuring the corresponding node identifier for each sub-test data; or/and, referring to the data in the test script The step of replacing with the sub-test data corresponding to each node further includes: configuring the corresponding node identifier for the test script corresponding to each node.

Further, the step of pushing the test program, sub-test data and test scripts corresponding to each node to the nodes of each slave server includes: copying the test program to each node of the slave server; configuring the test program of each node The folder identified by the corresponding node; pushing the sub-test data and test script corresponding to each node to the corresponding node.

Further, the node is an XML node in XML format for the test script, and the distributed performance testing method further includes: using xmlstarlet to search for all XML nodes that reference test data files, and storing the referenced test data file names in an array , to obtain the file array; traverse the file array, and execute the step of replacing the data reference in the test script with the sub-test data corresponding to each node.

Further, the configuration information of each slave server is stored in a global array, and each value of the array represents the configuration information of a slave server.

Further, the test program is Jmeter, and the distributed performance testing method also includes after each node executes the task and feeds back the execution results: utilize shell script to merge the feedback results, call the plug-in of Jmeter to analyze and count the feedback results, and drawing.

Further, the distributed performance testing method also includes after each node executes the task and feeds back the execution result: the summary data of the test result is stored in a CSV file, and java script script is written, and the chart and the statistical CSV file are combined After parsing, load it into an HTML template for display.

A distributed performance testing device of the present invention, comprising: an analysis unit, configured to analyze and obtain configuration information of each slave server according to pre-stored node allocation information, the configuration information includes a network address and the number of nodes allocated by each slave server; configuration The unit is used to split the total test data according to the configuration information of each slave server, obtain the sub-test data corresponding to each node, and replace the data reference in the test script with the sub-test data corresponding to each node, and establish The test script corresponding to each node; the test program, the sub-test data corresponding to each node and the test script are pushed to the nodes of each slave server; the test control unit, according to the test task, sends corresponding commands to the nodes of each slave server, by Each node executes the task and feeds back the execution result.

Further, the node is an XML node in XML format used by the test script; the configuration unit includes the xmlstarlet tool, which is used to search for all XML nodes that quote the test data file, and saves the referenced test data file name in an array, The file array is obtained; the file array is traversed, and the step of replacing the data reference in the test script with the sub-test data corresponding to each node is executed.

Further, the distributed performance testing device also includes: a data processing unit, which is used to merge the feedback results by using the shell script, call the plug-in of Jmeter to analyze and count the feedback results, and draw a graph.

Further, the described distributed performance testing device also includes: a display unit, the summary data used for the test results is stored in a CSV file, and java script scripts are written, and the charts drawn and the statistical CSV files are parsed and then loaded into the displayed in the HTML template.

The present invention is a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned distributed performance testing method is realized.

Each embodiment of the present invention automatically splits a test task into multiple nodes and performs synchronously to meet the requirements of a large amount of concurrent test scenarios; reduce labor costs, reduce working time and location and other conditions; reduce manual errors in data clipping/script editing ,reduce risk.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate the embodiments of the invention and together with the description serve to explain the principles of the invention. In the drawings, like reference numerals are used to denote like elements. The drawings in the following description are some, but not all, embodiments of the present invention. Those skilled in the art can obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flow chart of a distributed performance testing method provided by an embodiment of the present invention;

Fig. 2 is the flow chart of another kind of distributed performance testing method provided by the embodiment of the present invention;

Fig. 3 is a structural block diagram of a distributed performance testing device provided by an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined arbitrarily with each other.

The distributed performance testing method, device and computer-readable storage medium involved in the implementation of the present invention will be described in detail below in conjunction with the accompanying drawings.

Embodiment one:

Referring to Fig. 1, a distributed performance testing method includes:

First, according to the pre-stored node allocation information, the configuration information of each slave server is obtained by parsing, and the configuration information includes the network address and the number of nodes allocated by each slave server;

Secondly, according to the configuration information of each slave server, the total test data is split to obtain the sub-test data corresponding to each node, and the data reference in the test script is replaced with the sub-test data corresponding to each node, and each node is established. Corresponding test scripts; push the test program, sub-test data and test scripts corresponding to each node to each node of the slave server;

Wherein, the step of pushing the test program, the sub-test data corresponding to each node and the test script to the nodes of each slave server may specifically include:

The test program is copied to each node from the server;

The test program of each node is configured with the folder corresponding to the node identification;

Push the sub-test data and test script corresponding to each node to the corresponding node.

Again, according to the test task, corresponding commands are sent to the nodes of the slave servers, and each node executes the task and feeds back the execution result.

Specifically, the step of splitting the total test data to obtain sub-test data corresponding to each node may further include: configuring a corresponding node identifier for each sub-test data. The step of replacing the data reference in the test script with sub-test data corresponding to each node may further include: configuring the corresponding node identifier for the test script corresponding to each node.

In addition, the configuration information of each slave server can be stored in a global array, and each value of the array represents the configuration information of a slave server.

Through this embodiment, a test task can be automatically split and performed synchronously on multiple nodes to meet the requirements of large-scale concurrent test scenarios; reduce labor costs, reduce working time and location and other conditions; reduce labor in data clipping/script editing Mistakes, reduce risk.

Embodiment two:

Referring to Fig. 2, a distributed performance testing method includes:

Step 201: Split test data files;

Specifically: the node allocation information is stored in a file, the shell script reads this file, and saves it in a global array after parsing. Each value of the array represents the configuration information of a server, including IP and the number of allocated nodes, and calculates the total number of nodes. Then we split the test data into multiple parts on average, and append a mark after each, indicating which node to use.

Step 203: replace the test script;

Specifically: use xmlstarlet to search for all XML nodes that reference test data files, and save the referenced test data file names in an array. Traverse the file array, replace the corresponding data reference in the Jmeter test script, replace it with the test data file split in the previous step, save it as multiple copies of the test script, and append a mark (mark and One-to-one correspondence with the test data files in the previous step), this mark indicates which node to use.

Step 205: Deploy pressure nodes;

Specifically: traverse the global array of node allocation information, use the scp command to copy the local Jmeter program to the nodes of each server, and add a tag to the folder name, which is split from the previous test data file and replaced by the test script The marks in are in one-to-one correspondence.

Step 207: push the test script;

Specifically: Traverse the global array of node allocation information, use the scp command to distribute the edited test script copy and the split test data file to each node

Step 209: manage the control node;

Specifically: when executing the test task, the script traverses all node arrays, sends corresponding commands through the ssh command, and controls (start/stop) each node respectively; after the program starts successfully, the corresponding Jmeter node will appear in each server thread Information, we can use ps, grep and other commands to query and filter to determine whether the startup is successful. At the same time, it monitors the running status of the program through the unique information in the Jmeter thread. These can all be achieved through shell scripts on the master server.

Step 211: collecting results, statistical drawing;

Specifically: when each node executes the test, it saves the original data of the test result in a designated directory. After all nodes are tested, these results are recalled from the master server. Then use the shell script to merge these data, call the plug-in of Jmeter to analyze and count the data, and draw the graph. The summary data of the test is saved in a CSV file, and the files and graphs are saved in a unique directory.

Step 213: writing a test report;

Specifically: write a javascript script, parse the statistical CSV file and load it into an HTML template for display, and at the same time load the drawn chart into the template for testers to view.

This embodiment sets the test environment by self-definition, automatically recognizes the needs of the test environment by the script, and distributes the Jmeter program to the specified server; the test data script is automatically split, the test script automatically searches for keywords, and replaces the reference of the data file; During task execution, it is controllable in real time, and the running status is reported at any time; the test results are automatically collected and merged, and then the test results are counted, and an intuitive and readable test report is output.

Embodiment three:

Referring to shown in Figure 3, a distributed performance testing device, comprising:

The parsing unit is used to parse and obtain the configuration information of each slave server according to the pre-stored node allocation information, and the configuration information includes the network address and the number of nodes allocated by each slave server;

The configuration unit is used to split the total test data according to the configuration information of each slave server, obtain the sub-test data corresponding to each node, and replace the data reference in the test script with the sub-test data corresponding to each node, Establish test scripts corresponding to each node; push the test program, sub-test data and test scripts corresponding to each node to each node of the slave server;

The test control unit sends corresponding commands to the nodes of the slave servers according to the test tasks, and each node executes the tasks and feeds back the execution results.

During specific operations, the node is an XML node in XML format for the test script;

Described configuration unit comprises xmlstarlet tool, is used for searching the XML node of all referenced test data files, saves the referenced test data file name in an array, obtains file array; Traverses file array, and executes described will test script The data reference of is replaced by the sub-test data corresponding to each node.

Preferably, the distributed performance testing device further includes: a data processing unit, which is used to merge the feedback results by using shell scripts, call Jmeter plug-ins to analyze and make statistics on the feedback results, and draw graphs.

Preferably, the distributed performance testing device also includes: a display unit, the summary data used for test results is stored in a CSV file, and a java script script is written to parse the drawn chart and the statistical CSV file and load it into displayed in the HTML template.

The principle and working process of the above-mentioned distributed performance testing device are briefly described as follows: it is realized by shell script on the Linux platform. Shell scripts are simple and easy to use, suitable for processing objects such as files and directories, and can quickly process some very complicated things in a simple way. We take advantage of the flexibility of the shell and combine the small tools under the Linux platform to realize the intention of the scheme.

First of all, in view of the fact that Jmeter's test script is a file in XML format, we use the xmlstarlet tool to edit the script. As for the splitting of data files, tools such as wc, head, and tail that come with the Linux platform can be flexibly combined to complete work tasks.

Secondly, for tasks such as copying/distributing files, the Linux platform has opened the SSH service, and we can use scp to implement the deployment of the test environment, as well as the distribution of test plan scripts and test data files. For the client's control of each node (start/stop/monitor), we still use the SSH service to send relevant commands to each node to realize the management of Jmeter nodes.

Finally, we use Jmeter plug-ins to complete the collection of results and the writing of test reports. This is a set of plugins that are independent of Apache Jmeter and provide customization. It has excellent drawing, loading methods, and also provides a richer function library. We combine the shell script, quote the jar package of the plug-in, integrate the collected result data, count and draw, and finally present the statistical file in a templated HTML file.

There is no manual intervention in the whole process of this embodiment, which can reduce labor intensity, save labor and time costs, and avoid work accidents introduced by manual editing; multiple Jmeter instances can be run on one server, greatly reducing the requirements of the test environment; each Jmeter instance Running independently, there is no data synchronization during the test run, and each node is flat, which avoids the problem of overloading the system overhead of the Jmeter client in the scenario of multiple remote nodes, and reduces the risk of contamination of test results.

The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned distributed performance testing method is provided. It has the corresponding technical effect of the test method, which will not be repeated here.

Those of ordinary skill in the art can understand that all or part of the steps/units/modules of the above embodiments can be implemented by hardware related to program instructions, and the aforementioned programs can be stored in a computer-readable storage medium. When the program is executed, Execution includes the corresponding steps in the units of the above-mentioned embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (12)

1. A distributed performance testing method, characterized in that, comprising: According to the pre-stored node allocation information, the configuration information of each slave server is obtained by parsing, and the configuration information includes the network address and the number of nodes allocated by each slave server; According to the configuration information of each slave server, the total test data is split to obtain the sub-test data corresponding to each node, and the data reference in the test script is replaced with the sub-test data corresponding to each node, and the corresponding sub-test data of each node is established. Test script; push the test program, sub-test data and test script corresponding to each node to each node of the slave server; According to the test task, corresponding commands are sent to the nodes of the slave servers, and each node executes the task and feeds back the execution result. 2. the distributed performance testing method as claimed in claim 1, is characterized in that, The step of splitting the total test data to obtain the sub-test data corresponding to each node also includes: Configure the corresponding node identifier for each sub-test data; or/and, The step of replacing the data reference in the test script with the sub-test data corresponding to each node also includes: The corresponding node identifier is configured for the test script corresponding to each node. 3. the distributed performance testing method as claimed in claim 1, is characterized in that, The step of pushing the test program, the sub-test data corresponding to each node and the test script to the nodes of each slave server includes: The test program is copied to each node from the server; The test program of each node is configured with the folder corresponding to the node identification; Push the sub-test data and test script corresponding to each node to the corresponding node. 4. the distributed performance testing method as claimed in claim 1, is characterized in that, described node adopts the XML node of XML format for test script, and described distributed performance testing method also comprises: Use xmlstarlet to search for all XML nodes that reference test data files, save the referenced test data file names in an array, and obtain a file array; Traverse the file array, and execute the step of replacing the data reference in the test script with the sub-test data corresponding to each node. 5. The distributed performance testing method according to claim 1, wherein the configuration information of each slave server is stored in a global array, and each value of the array represents a slave server configuration information. 6. the distributed performance testing method as described in any one in claim 1-5, it is characterized in that, described test program is Jmeter, and described distributed performance testing method also returns after each node performs task and feedback execution result include: Use the shell script to combine the feedback results, call the Jmeter plug-in to analyze the feedback results, and draw the graph. 7. the distributed performance testing method as claimed in claim 6, is characterized in that, described distributed performance testing method also comprises after each node performs task and feedback execution result: The summary data of the test results is saved in a CSV file, and a java script script is written to parse the drawn chart and statistical CSV file and load it into an HTML template for display. 8. A distributed performance testing device, characterized in that, comprising: The parsing unit is configured to parse and obtain the configuration information of each slave server according to the pre-stored node allocation information, the configuration information including the network address and the number of nodes allocated by each slave server; The configuration unit is used to split the total test data according to the configuration information of each slave server, obtain the sub-test data corresponding to each node, and replace the data reference in the test script with the sub-test data corresponding to each node, Establish test scripts corresponding to each node; push the test program, sub-test data and test scripts corresponding to each node to each node of the slave server; The test control unit sends corresponding commands to the nodes of the slave servers according to the test tasks, and each node executes the tasks and feeds back the execution results. 9. the distributed performance testing device as claimed in claim 8, is characterized in that, described node adopts the XML node of XML format for test script; Described configuration unit comprises xmlstarlet tool, is used for searching the XML node of all referenced test data files, saves the referenced test data file name in an array, obtains file array; Traverses file array, and executes described will test script The data reference of is replaced by the sub-test data corresponding to each node. 10. The distributed performance testing device according to claim 8 or 9, further comprising: The data processing unit is used for merging the feedback results by using the shell script, calling the plug-in of Jmeter to analyze and count the feedback results, and drawing. 11. The distributed performance testing device according to claim 10, further comprising: The display unit is used to save the summary data of the test results in a CSV file, write javascript scripts, parse the drawn charts and statistical CSV files, and load them into the HTML template for display. 12. A computer-readable storage medium, on which a computer program is stored, wherein when the program is executed by a processor, the distributed performance testing method according to any one of claims 1-7 is implemented.