Forum Discussion

jkrolczy's avatar
jkrolczy
Regular Contributor
4 years ago

Distributed Testing: Unknown ??? is causing Master to stop all Slaves executing scripts

I have a case open on this, but I am querying the TestComplete community as well.

 

Distributed Testing ( Yes it is deprecated ... known)

 

But there are still users out there, like myself that rely heavily on this until a migration to, in my case, Azure DevOps.

 

I am querying the TestComplete users to see if this has been seen by anyone and possible areas to look into for debugging and correcting.

 

I have Distributed Testing setup for a Master to 5 Slaves.

I use a Win10 Hyper-V setup.

TC\TE 14.71   ... tried 14.80 also

 

All of my Distributed Testing has been running fine until lately.

  Microsoft patch affecting me ?

  Hyper-V cluster\node affecting me ?

  Networking issue ?

 

This issue being seen is when I have started my Distributed Testing

either from a Team City Job or Manually from the Master, during the

scripts being executed by the Slaves there is some unknown ???

that causes the Master to tell all the Slaves to stop all execution.

 

Looking at the logs from all the Slaves, the stop is done at the same timestamp.

 

I am NOT manually doing a stop from the Master.

 

This issue is frustrating because I once receiving completed script executions before, but now

I get Distributed testing runs that are incomplete runs missing scripts at the end.

 

This happens with a run of 12 scripts or runs with over 100 scripts and the stop comes at random times.

 

I am trying to figure out a way to monitor the port that the Master uses talks to the Slaves to see what\might something is sent to stop them.  Not sure how to do this just yet.  Suggestions ?

 

Thanks

 

 

 

 

Distributed Testing
  • vinniew's avatar
    vinniew
    Staff
    4 years ago

    Hi jkrolczy 

     

    Hope you are well. 


    Are the errors you get when kicking off manually and Team City the same? 

     

    Are you able to share the console logs? 

     

    This may useful for community members to see and then advise

     

    I am not well versed in this but it would be interesting, hearing other views on this matter

     

    KR

    Vinnie

    • jkrolczy's avatar
      jkrolczy
      Regular Contributor
      4 years ago

      I am at the point with this issue thinking that a powershell script is being executued on my Master during a Distributed run.  This powershell script is from the Team City side and may be a script trying to take down the job due to something Team City detected.

       

      I am testing out a ADO pipeline doing a Distributed run to the same slaves.

      More info to come after this testing is done.

      • jkrolczy's avatar
        jkrolczy
        Regular Contributor
        4 years ago

        Current Status as of June 10, 2021

         

        Running a Team City job that kicks off Distributed Testing and now using an ADO pipeline to kick off the same commandline

        for Distributed Testing, the issue of the script run ending\stopping at some random time is still occurring.

         

        My suspensions now is maybe how Hyper-V works ?

        Session load balancing, networking connection break, or something else that could be affecting the Master and somehow killing my Slaves running scripts.

        Some anomaly is affecting my Master to trigger a stop to my Slaves and I cannot figure this one out still.

         

        I was hoping the issue was more Team City related, but that is not the case now.