IVI drivers cleaning after TestStand caller application crashes

[sorinxxxx]

Hi everybody

One of our clients is using NI TestStand and calling a LabWindows/CVI application that calls a custom IVI driver.

Sometimes the TestStand application crashes and as result the IVI driver remains in memory and next time when user wants to restart the TestStand application the IVI driver would prevent that. At least this is the custmer opinion about failing to restart the TestStand application once it crashed and the only way to restart the TestStand application is to reboot the PC.

Would anyone had this kind of interaction between TestStand <-> LabWindows/CVI <-> IVI driver where crsahing TestStand once force the user to reboot his PC?

Thanks

Sorin

Message Edited by sorinvalea on 11-03-2009 04:43 PM

[Kelly_R]

Hi Sorin,

We will need a little bit more information from your client to help debug why this crash is occuring.

1. What version of TestStand and CVI are you using?
2. What do you mean by TestStand crashing?  Does it hang, give an error message, etc?
3. How often does the crash occur (e.g. everytime the sequence is run, once a week, etc)?
4. Can you run the CVI code module that calls the custom IVI driver itself without error? If you are only seeing the error occasionally, we may need to test this repeatedly.
5. Can you get an NI Spy capture for me during a session when the crash occurs?  This will help us narrow down exactly where in the code module the crash is occuring. Here is a link with more information on how to use NI Spy.
6. What happens when you try to restart TestStand?

[sorinxxxx]

Hi

Thanks for your reply. Unfortunately at this time I could not ask the customer for these specific details as someone else from customer service must go there and get more info about the system.

I thought maybe someone else experienced or hear about this particular issue regarding TestStans->LabWiindows/CVI->IVI driver where the crashing in the TestStand would force the user to reboot the PC because of the IVI driver left in memory.

Sorin

[Kelly_R]

Hi Sorin,

It does sound like there is information that is stuck in memory that is causing this issue.  When the TestStand process is killed due to the crash, certain DLLs, etc may be orphaned as there was not a proper clean up.  Restarting the computer will clear out this memory.  This is even a bigger problem since we are working with hardware here.  We can always open up new software resources, but we can only open up the one hardware resource.  Currently this hardware resource is stuck in memory, which is most likely why we cannot launch our sequence.  We do not have any known issues with this combination.  My best recommendation is to get in touch with your customer so we can figure out exactly what is causing this crash. Unfortunately there is not a lot we can do without this additional information. 

[sorinxxxx]

Hi

I want to know more about the "hardware resource" that could open only one time in memory. What is considered a "hardware resource" by TestStand? How it handles this "hardware resource" when has been left in memory from a previous crashed session?

I will come back later when I will have more specific information about this case.

Thanks for feedback

Sorin

[Kelly_R]

Hi Sorin,

This is somewhat of an oversimplification, but you can think of this situation kind of like a telephone.  If I am speaking with you on the phone, but all of a sudden I fall walk away, you are still waiting for communicate with you, and if someone else calls and tries to make a new connection, they find the line is busy.  This parallels the hardware resource conflict that we are seeing.  We can only have the one reference to the hardware (what you are talking to with the IVI driver), and when TestStand crashes, this resource is left hanging because things were not shut down properly so we cannot open up a new line of communication with it.  All of this is on the driver level, so you would see the same behavior in areas other than TestStand as well.  Once the crash has occurred, we orphan this hardware resource, and again, need to restart to reset everything.  

Please let us know when you have additional information, and we can continue to troubleshoot.  My guess is there is something that is throwing an exception within the custom driver or the CVI code.

[sorinxxxx]

Hi Kelly

Thanks for your extra information regarding this subject. It was very useful.

I will come back with specific details about this crash as soon as I will get them.

Sorin