Performance Test Crashes

[Ian]

Hello, for some reason my performance tests have been causing Telerik to crash about half of the time that I run them. I haven't had this problem before today.

What happens is that I run a performance test, it completes, then while loading the result the application stalls for a second, then closes itself. Upon restarting I can see the test result stored in the app (sometimes with a clearly inaccurate runtime), but any attempt to open it crashes the app. What's going on?

[Ian]

After some investigation it seems that the crashes are somehow related to whether or not the script ends on a verification step.

[Cody]

Hello Ian,

I am sorry you are running into this problem. I see quite a few crash reports from you logged in our system. They all say the same thing:

USER: [e-mail]
MACHINE: VIC-GFPM4S1
PRODUCT: TESTSTUDIO
VERSION: 2014.4.1411.0
WINDOWS: Windows 7
TIME(local): 2015-03-23 8:51:14 AM
.NET Version: 4.0.30319.18063
Current Culture: en-US
OS Culture: en-US
Working Set Memory:375009280
GC Heap Memory:175918912
ERROR: System.ArgumentException Width and Height must be non-negative.
Width and Height must be non-negative.
at Telerik.TestingFramework.Wpf.RatioCanvas.ArrangeOverride(Size arrangeSize) at System.Windows.FrameworkElement.ArrangeCore(Rect finalRect) at System.Windows.UIElement.Arrange(Rect finalRect) at System.Windows.ContextLayoutManager.UpdateLayout() at System.Windows.ContextLayoutManager.UpdateLayoutCallback(Object arg) at System.Windows.Media.MediaContext.InvokeOnRenderCallback.DoWork() at System.Windows.Media.MediaContext.FireInvokeOnRenderCallbacks() at System.Windows.Media.MediaContext.RenderMessageHandlerCore(Object resizedCompositionTarget) at System.Windows.Media.MediaContext.RenderMessageHandler(Object resizedCompositionTarget) at System.Windows.Threading.ExceptionWrapper.InternalRealCall(Delegate callback, Object args, Int32 numArgs) at MS.Internal.Threading.ExceptionFilterHelper.TryCatchWhen(Object source, Delegate method, Object args, Int32 numArgs, Delegate catchHandler)

This appears to be some sort of a UI bug. Last week we released a new major version, 2015.1.319, with a significant change to the UI. The crash report indicates you're running 2014.4.1411. Can you upgrade to this new version and verify whether or not this problem magically was fixed? You can download it directly from this link after logging in to Telerik.com.


Regards,
Cody
Telerik

 

[Ian]

It appears to have magically been fixed by the update, thanks Cody!

[Ian]

I was wrong, its still failing in the same way with some regularity.

[Cody]

Hi Ian,

All right. Can you create a test that uses any publicly accessible web site I can run on my machine to reproduce the problem? Once I can reproduce the problem I'll file a problem report with our development team.


Regards,
Cody
Telerik

 

[Ian]

I reproduced it on Amazon.com:

https://drive.google.com/file/d/0ByqocPEYOq72VDlqajdBRnQyLUk/view?usp=sharing

[Ian]

The link is for the .tstest file

[Ian]

Just realized that this forum has an option to upload files, I feel silly.

[Ian]

Here is the file

[Ian]

okay the file upload is kinda buggy let's try that again

[Cody]

Hi Ian,

Just realized that this forum has an option to upload files...

Yep it does. Just watch out we only accept file types zip, rar, ace, jpg, gif, css, png. Many people try to upload other file types such as .txt or .docx and then wonder why it didn't work. You got it figured out however.

I have tried to reproduce your crash using your sample test over a dozen times... no luck. Does the speed of the machine make a difference? Fast or slow machine? Does it matter which browser you use? IE versus Chrome versus Firefox? Does the version of Windows make a difference? I'm using Windows 7 64-bit. Does it make a difference if the verify step normally passes or normally fails? What if you change the Verify step into a Wait For step as shown in the attached screen shot?


Regards,
Cody
Telerik

 

[Ian]

The crash is not consistent, sometimes it will crash with every test, other times it will rarely crash. I am running windows 7 64-bit and I have encountered this bug in firefox and chrome. It seems like the bug isn't triggered by the test itself, but when Telerik tries to read the performance data from the test. The test will complete and then Telerik will crash as soon as the results load. If I open up Telerik again and try to view that same result Telerik will crash.

If you try opening the first test in the performance test history of the 'repro' test I sent you (firefox, 2 seconds long), Test Studio should crash.

[Cody]

Hello Ian,

What you sent me was the test only. The results are stored as .tsperf files in the folder you configured for Performance testing to store the results to - see attached screen shot for an example. If you can locate your results I'll be happy to load them up here and see what happens.

Your symptom description makes it sound like the crash can happen by just opening the results from History view. Is this true? I actually do hope so because that would be a much easier and more reliable repro.


Regards,
Cody
Telerik

 

[Ian]

Yes, this crash can be repro'd from the history view. I have attached the test results.

[Cody]

Hello Ian,

Excellent! I was able to repro a crash with the result files. I filed a problem report on it here. I've requested development to see if they're able to escalate this one (no promises yet).


Regards,
Cody
Telerik

 

[Cody]

Hi Ian,

I have a little bit more good news. Our development team has confirmed this is a bug in Test Studio and has an estimate of two weeks to have a fix ready.

Regards,
Cody
Telerik

 

[Cody]

Hi Ian,

I have been corrected on our scheduling. Your bug is scheduled to be worked on in the next two weeks, but we do not have a release currently planned until sometime in May. We might have an internal build between now and then, but it hasn't been officially decided.


Regards,
Cody
Telerik

 

[Ian]

Good to hear, thanks Cody

[Cody]

Hi,

You're welcome. Keep monitoring the problem report. Development will update it when the fix has been found and applied. This ticket will automatically close after 7 days of no communication activity in it.


Regards,
Cody
Telerik

 

[Cody]

Hello,

In case you didn't get notice, our developers have fixed the cause of the crash today. The fix will be included in our next release which is due out later this month.

Regards,
Cody
Telerik

 

[Ian]

I saw that, thank you.