locked
capturing test results RRS feed

  • Question

  •  

    I am all-around new to team testing, so please forgive my stupidity. 

     

    The questions: 

     

    1) What is the best way to access test results programmatically so that we can collect access/response time data and mine it?

     

    2) The tests generate a huge binary file (I guess test results?).  What is this file?  How can I use it?

     

    3)  Is there a way to set the tests to run daily at a specific time?

     

     

    We are testing an application's performance with 2100 hits from three different cities, and we're trying to analyze the results. 

     

    As of right now we're doing it quick and dirty - we copy all rows into a text file and run a parser on it and generate a comma-delimited file for excel.  From there we run our statistics on it. 

     

    I'm sure there's a better way to access the results, but right now we're just sticking to what works.

     

     

    Friday, October 19, 2007 2:26 PM

Answers

  • I assume you are talking about analyzing load testing results.

     

    The load test result analyzer would be the place to view the results for a load test.   A test run will generate a .trx file that contains test result information.  In the case of a load test, the .trx does not have the response time information, but it includes the information needed to attach to the load test result storage.

     

    If you open the .trx file, VS will open and load the result in the "Test Results" window.  From there you can choose to "View Result Details" which will open the load test result analyzer.

     

    Here's a document that describes the load test analyzer...

    http://msdn2.microsoft.com/en-us/library/aa730850(vs.80).aspx

     

    2) I assume the huge binary file is the .trx file.  The .trx contains the test result.  In the case of a load test though, peformance counter data is contained in the load test result store (the .trx helps to locate the results store and specific run).

     

    3) I can can use the command line utility "MSTEST.EXE" to start a test from a script.

     

    Hope this helps,

    Rick

    Friday, October 19, 2007 3:09 PM
    Moderator
  • A custom Request-Plugin provides a PostRequest method that makes the entire response available, from which you could find the data that is reported in the web test result viewer.

     

    I can see now that you are talking about a web test result.  I'm curious to know why the test result is so large.  Are you running a large number of iterations of this web test?   Would it help to include the web test in a load test, run the multiple iterations with the load test, and have the results stored in the load test result repository?

     

    Thanks,

    Rick

    Monday, October 22, 2007 2:49 PM
    Moderator

All replies

  • I assume you are talking about analyzing load testing results.

     

    The load test result analyzer would be the place to view the results for a load test.   A test run will generate a .trx file that contains test result information.  In the case of a load test, the .trx does not have the response time information, but it includes the information needed to attach to the load test result storage.

     

    If you open the .trx file, VS will open and load the result in the "Test Results" window.  From there you can choose to "View Result Details" which will open the load test result analyzer.

     

    Here's a document that describes the load test analyzer...

    http://msdn2.microsoft.com/en-us/library/aa730850(vs.80).aspx

     

    2) I assume the huge binary file is the .trx file.  The .trx contains the test result.  In the case of a load test though, peformance counter data is contained in the load test result store (the .trx helps to locate the results store and specific run).

     

    3) I can can use the command line utility "MSTEST.EXE" to start a test from a script.

     

    Hope this helps,

    Rick

    Friday, October 19, 2007 3:09 PM
    Moderator
  • Thanks Rick, I do have a few more questions.

     

    The test results files generated are close to 3GB apiece and they kill Kenny when we double click them.  Therefore what we do is a R-Click, "Copy all Rows" from the webtest window, paste in a text file, then we run a custom-parser on it to get the data we need.

     

    I saw the MyClassCleanup and MyTestCleanup methods which run at the end of a successful test and would like to use them to automate our existing system. 

     

    How can I access the text-form of test results as they occur?  The idea is to get the text-form results of each test run and append them to a text file, then at the end of the test run the parser.

    Friday, October 19, 2007 5:29 PM
  • A custom Request-Plugin provides a PostRequest method that makes the entire response available, from which you could find the data that is reported in the web test result viewer.

     

    I can see now that you are talking about a web test result.  I'm curious to know why the test result is so large.  Are you running a large number of iterations of this web test?   Would it help to include the web test in a load test, run the multiple iterations with the load test, and have the results stored in the load test result repository?

     

    Thanks,

    Rick

    Monday, October 22, 2007 2:49 PM
    Moderator