Shulou(Shulou.com)06/02 Report--

Catalogue specification

In order to develop a good habit of performance testing, we need to do an important job before starting LoadRunner: create a new directory.

Create a folder under the root directory of disk D (whichever disk you want) and name it "performance Test". This folder is retained for a long time and is used to standardize the management of related files involved in performance testing. In the performance Test file, create a new project folder using the name of the project + performance test start date. For example: "car talent 20141111". In the project folder, create four new subdirectories, namely: script, scenario, result, and data, as shown in the figure:

1. Script directory

We run Virtual User Generator, record and write the script, and store the script in this directory. The naming of the script file should follow the key business behavior that describes the script. Such as: login, registration

2. Scene directory

After we run Controller, load the relevant scripts, and set the scene policy, we save the scene in this directory. Just run it directly next time. The naming method of the scene file is mainly to describe the business, and there is no need to mark the number of concurrent people in the scene and other information.

3. Result catalog

In Controller, set the location of the result file for a scenario to run for later test analysis. The saving and naming rules of the result file are particularly important to the process of performance test and analysis. In the Controller interface, go to the menu "Results" > > "Results Settings..." To set the relevant properties for saving the result file, as shown below:

[Results Name] enter the number of VU for this scenario after the default res letter, and then enter the business scenario description. For example: "res100 people visit Auto Talent Home Page"

[Directory] Select the result file storage directory, and directly select the "results" directory that we have prepared.

[check box] Select the red coil labeling option in the figure, meaning: there is no need to confirm to directly overwrite the result file with the same name. In practical work, for example, if we run a 100-person login scenario, we may run multiple times, so the result file can only keep the last result. However, when we change the number of VU settings for the scene to 200people, we must change the result name Results Name again, changing "res100 login" to "res200 people visit the home page of car talent", so run the scene again and generate a new result file.

4. Data catalog

In this test project, all the parameterization involved in the script and the required data source files (txt) are saved in this directory. This work should be well prepared before opening the LoadRunner tool.

Data source preparation includes:

(1) understand the business scenario, design which business data needs to be parameterized, and how much quantity is needed. It is exported from the database or generated by Excel, and finally pasted into the txt file.

(2) analyze whether there is a necessary corresponding relationship among the parameters. For example: user name and password, if these are two parameters that need to be parameterized. Then each value must ensure that the two parameters are the same row of data. In such cases, all parameters with corresponding relations must be reflected in a txt file, with each column of data corresponding to each parameter, and the row position corresponds to the correct position.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.