For more information, see the BizTalk hands-on experiment series directory.
BizTalk Development Series
Performance Testing and Analysis of BizTalk applications is a very important process, bizTalk applications are greatly affected by the deployment structure, number of message requests, and Message Size During Run-time, therefore, both simple and complex applications need to be executed before being deployed to the production environment. In performance testing, the Performance Monitor provided by the Windowns system is used to collect relevant data and determine the performance bottleneck and load size of applications based on the value of these counters.
Although you should be familiar with Windows Performance Monitor, if the usage method is incorrect, you may not only not be able to get the data you want, but may also be able to perform a test in vain. Next, we will discuss how to use the windows performance monitoring tool in the BizTalk performance test.
The reliability and performance monitor of Windows2008 provides more user-friendly functions (such as) in data reporting and dataset management ). This demo is based on a performance monitor and consists of three parts: data collection, data viewing, and data analysis.
Create a Data Collector set to collect data
1. Click create Data Collector set in Data Collector Sets> User-Defined.
2. manually create a data collector set
3. Types of data to be collected in the dataset
4. Add corresponding counters. Because different performance comparisons require different data, we recommend that you select all counters under the corresponding instance.
6. No registry key is recorded in "which registry key do you want to record"
7. Select the default storage location
8. In "whether to create a data collector set", select Save and disable
9.Enable report generationRight-click the new Data collector BTS_Demo. Select Data Manager and click "enable Data management and report generation" in Data management ". This step is important. If this item is not enabled, the latest report feature of the Data Collector will not be available.
10. OK. The data collector set has been created. Enable the new data collector BTS_Demo before performing performance tests on BizTalk.
View performance data
View current performance dataYou can view counter changes in real time on the current activity page of the Performance Monitor. Add counters as needed before use.
View performance Log DataIn the performance monitor window, click View log data. In the displayed window, select the location and time range of the log file to be tracked. By viewing the log data, you can view the data changes throughout the test.
View performance reports and analysis
For test data, you generally pay attention to the maximum value, minimum value, average value, test time, system configuration, and other information. With the data collector created earlier, you only need to click the latest report function of the data collector. You can view the generated tracing report in the report. The following report shows the numerical statistics of each counter.
PAL analysis tools
PAL is a performance analysis tool on Codeplex. It provides log analysis templates for multiple application systems and can be customized to create templates.
After you specify the log file to be analyzed and the analysis template in PAL, PAL will generate HTML-type reports. Each counter has detailed statistics on the image, if some values exceed the regular range, a color reminder is displayed. Unfortunately, the tool does not support Chinese characters.
Other Tips
1. Save the counter
In the counter list in the performance Viewer window, right-click and select "Save settings ". You can save the settings as HTML. You only need to drag the HTML to the counter list during use.
2. Save the data collector set settings
Right-click the data collector and choose Save template to save the data collector set. Use the template to create a data collector set.
Summary
Through the above demonstration, the usage of Performance Monitor in BizTalk performance testing is basically described, in actual use, add more counters when adding counters in order to make the tracking data meet different performance comparisons.