09-15-2008 11:04 AM
An operation is getting the following error when attempting to make a NISE connection. He is running a sequence in TestStand 4.0 and using the switching properties of the step. This is the first switching operation in the test sequence.
An error occurred while performing the switch operation 'Connect' for the following device '265869_T45_LED_Driver'. Could not perform the switching operation: Details: (IVI Error Code: BFFA1190) [Device: NI_PXI-2532_ivi](Error = -1074130544) [Error Code: -17500, Operation Failed.]
I cannot find any information on any of the above error codes. What exactly does this error relate to?
I am using TestStand 4.0, and NISE 2.11
09-15-2008 01:03 PM
IVI Error Code: BFFA1190 (hex notation for -1074130544) is thrown when there is an invalid session to the switch device. This error occurs when a session cannot be opened to the instrument. Error Code: -17500 is a generic Internal Error in TestStand.
Do you have LabVIEW installed? If so, could you attempt running an example installed with NISE, such as NI Switch Executive - Getting Started.vi to verify this also returns a similar error?
One problem that could cause this error is inability to communicate with the hardware. Can you open and close relays on this NI PXI-2532 using the Switch Soft Front Panel?
Start >> All Programs >> National Instruments >> NI-SWITCH >> NI-SWITCH Soft Front Panel
Also, verify the Virtual Device configuration validates and the switching operation works as intended in the NI Switch Executive Test Panel, available from within Measurement and Automation Explorer (MAX).
Let me know how it goes. Thanks!
Switch Product Support Engineer
NI - USA
09-15-2008 01:31 PM
Opening two sessions to the same device could also cause an invalid session error, however TestStand should handle the switch session management. One more question: Are you only performing switching operations using TestStand IVI Switch or NISE Steps, or is it possible you're opening an additional session to this NI PXI-2532 through an external code module (LabVIEW, CVI, etc.)?
Switch Product Support Engineer
NI - USA
09-17-2008 04:27 PM
The problem seems to be due to TestStand improperly releasing the reference to the NI Switch Executive session. After running a test, that fails the operator can click on Execute-Restart from our custom User Interface. By utilizing the Restart option the Close Session is never called, hence the first Connect call to the NISE session gives an error. Since I suppose the reference is still open?
Running the test ( TestStand -> Single Pass ) and then closing the execution propertly closes the reference.
MainSequence Cleanup : The step below is always called
11961. niSE_DisconnectAll (265869_T45_LED_Driver)
Process ID: 0x00000E84 Thread ID: 0x00000E14
Start Time: 16:22:54.288 Call Duration 00:00:00.032
Status: 0 (VI_SUCCESS)
Only after closing the execution ( NOT by clicking on restart ) does the following step run.
11979. niSE_CloseSession (265869_T45_LED_Driver)
Process ID: 0x00000E84 Thread ID: 0x00000C34
Start Time: 16:25:10.337 Call Duration 00:00:00.000
Status: 0 (VI_SUCCESS)
So the workaround is to have the operator NOT run restart on this type of test. This type of test is one that uses inline TestStand Switching via Switch Executive.
Another reason I have to explain to people here at my company why NOT to use Switch Executive.. UGH.....
09-18-2008 05:55 PM
Thanks for your post back.
You are correct that in NI Switch Executive you need to close the session that you are currently using before moving on to open another. I feel that in your case a session is attempting to open when another session is open for the same switch. Is it possible to put a step in your sequence to ensure that your session is closed? I see that if a UUT fails you want to restart the test, so there could be an "IF" condition to make sure the session is closed before the next open session occurs. Another option that you can do is to change the settings of the sequence. If you go to Configure >> Station Options there are some options where you can select Run Cleanup when you have a On Run-Time Error. There is also an option to immediately go to clean up on a sequence failure. In the clean up step there could be a close NI Switch Executive reference to make sure that when the test restarts the open session will not fail on Open Session. I don't feel like the problem here lies with Switch Executive since a session was attempting to open before closing.
I do want to ensure that your error is cleared up. Do you have a support contract with NI. There is a team here to help with specific TestStand errors such as this. We do not normally do this on the forums but if your error continuous would it be ok to contact you offline? Have a great day Teds!
09-26-2008 10:56 AM
The error only occurs when Excecute-> Restart is selected, this implementation of restart is defined by NI. Support found the following:
After digging around some more it looks like this is actually a known bug that is scheduled to be fixed in a future version (CAR ID#: 46891). It's also mentioned in the CAR notes that this behavior only occurs when using real hardware because simulation allows the use of the hardware from 2 virtual devices at the same time. As a workaround, what we need to do is to set the the following property to Nothing assuming the name of the device is "SwitchExecutiveExample":
You have to do this for all executions using the device.
If that variable has colons then you will have to change it to this: RunState.Execution.AsPropertyObject.SetValInterface("TSIviStepTypes_AnchorName_For_SE::SwitchExecutiveExample",0,Nothing)
09-29-2008 02:49 PM
Thanks for the post back and the information. What you have found will of course be used to improve TestStand and thanks for working with us. Let us know if there is anything else we can do.
10-20-2008 09:43 AM
I also ran into this problem... as expected since I'm running TestStand 3.5 and NISE 2.10. I used the fix from this thread and now everything's fine.
Does anyone know what release will fix this issue?
10-21-2008 10:42 AM
Thanks for your post!
Glad to hear that the work around has you up and running! I can say for sure that this will be fixed in future versions however at this time we cannot release that information publicly as things are not finalized. You can always email or call NI directly to check on the release notes of the CAR mentioned above in this thread. We always detail in the release notes of every version what is fixed from the previous version. So keep any eye on the TestStand page of ni.com to see what versions are coming up! Thanks for your questions and good luck to all on your applications.
10-22-2008 02:18 PM
Hey kwseitz3 and Paul,
After doing some more testing on our end, we believe that this might be a separate bug and would thus warrant a separate CAR to be filed. Do you still have the NI Spy captures from: 1. using Reset and 2. using Single Pass? You also mentioned that it worked just fine when you ran your test without hardware so it would also be helpful if you could get us NI Spy captures for Reset and Single Pass for this scenario as well.
One last piece of the puzzle that would help us would be your Switch Executive and IVI configuration information. For Switch Executive you could just export your configuration from MAX and for IVI you can copy the IVI configuration store file. Both of these should be .XML files.
NISE Configuration: In MAX, right-click Switch Executive Virtual Device, select Export. Send this file.
IVI Configuration: Click on IVI Drivers to find the Master IVI Configuration Store Location. Send the IviConfigurationStore.xml file.
I went ahead and attached screenshots of where you can grab these files.
Do you think either of you could get us these files so we can further isolate the cause of this issue?