From 04:00 PM CDT – 08:00 PM CDT (09:00 PM UTC – 01:00 AM UTC) Tuesday, April 16, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

NI TestStand

cancel
Showing results for 
Search instead for 
Did you mean: 

TerminateAll problem: TestStand is reporting Socket thread as terminating and still running steps in Main block.

This seems to be related to https://forums.ni.com/t5/NI-TestStand/Some-Executions-Are-Not-Terminated-by-Using-Terminate-All-in/m...

 

Here I am using a simplified sequence. It runs four sockets using batch model. A sub-sequence call uses serial synchronization. When socket 1 comes to execution, it sends Engine.TerminateAll(). All sockets read the terminating status as first step of Main block using ActiveX call RunState.Execution.GetStates and record it to results .

 

Sometimes Sockets 2 and/or 3 do report terminating status as 2 (=Terminating), but still run all the steps in Main block.

 

I used TS 2017 SP1. I also had similar behavior in TestStand 2020, but could repeat it using this particular sequence. Sometimes I need to repeat the test 5 - 20 times, before this behavior happens. Also, if Windows is otherwise busy, e.g. updating, it seems that this behavior does not happen.

 

It seems, that if synchronization steps are used instead of synchronization of sequence calls, this does not happen, but haven't totally confirm it, since I do have an automated system to test this hundreds of times.

 

I will attach an example sequence and a report showing this behavior in Socket 2 results. (I am not sure why results are not in numerical order, but that is not an issue to me).

 

To me, it seems that either there is a bug in TestStand or Terminating is just a suggestion and not pre-emptive deterministic thing or that I have understood something wrong.

0 Kudos
Message 1 of 3
(755 Views)

Thanks for the report. It took a while, but I was finally able to reproduce this. It is on my list to investigate further soon.

0 Kudos
Message 2 of 3
(692 Views)

I made a bug report about this, and they say, it's going to fixed in a future release.

 

NI was not able to replicate this, if batch synchronization steps were used instead of setting synchronization in a sequence call. So that works as a temporary fix.

0 Kudos
Message 3 of 3
(567 Views)