Intermittent Error behavior Launching Chrome

[WarrenBowman]

Here is the Error "ERROR: Internal : Failed to perform step 3 in Navigate Stage 'Launch' on page 'Launch' - Could not identify process owning the current foreground window". When Launching Chrome this error will occur and the process will fail. If you open the VM remotely and keep it open on your desktop the error doesn't occur. If you return to the running the process from control center without opening the VM first it will return to this error. At some point, it somehow rights itself. And you can return to normal operations.

The issue is not consistent, but it's been happening ever other day and can take hours to for it to stop. Since this is happening in a prod environment that runs 24\7 it's a major concern.

[david.l.morris]

Is there a chance the screen is locking?

[WarrenBowman]

Maybe locking, We've got a chain of processes running one after another, the last of which is set to run in exclusive mode. The thing is that this chain had been running for more than a year. On Jan 2nd or 3rd a group of our prod VMs went offline and we couldn't restore them and have them hold a "logged out" state. But we could get them to "idle" which meant leaving the VM running. The process in question runs every 30 mins on a specific VM and the error above started to occur, after we had modified our schedules to no longer login and out.  Initially I thought that might be the issue. But I'm not so sure. At some point, a tech fixed the logged out state problem. But we've continued to operate from idle. I'm starting to believe that this error is related to the chained processes not releasing fully before the next process runs. Plus, the fact that we have the process set to run exclusive mode. But it's running now and hasn't had the error today. I added so wait time between processes in the chain in scheduler. But the error has occurred 3 times before and when it does it's a problem for hours. One thing though. If you open the VM on your desktop and run the process while the VM is open it works.

[michaeloneil]

Hi @WarrenBowman 

Ive had a similar issue with a machine previously and it was caused by persistent connections after people had accessed the machine remotely. It resulted in more than one connection existing on the machine and the process failing to identify which connection was performing the action. Im guessing you leave the machines signed in while inactive? I would suggest setting up regular restarts of the machine to remove any persistent connections or sign out the machine while its not in use.

One way to test if this is the issue is, when the failure occurs in the process, open a command prompt from your machine and run this script qwinsta /server:{IP address of the resource machine} 

It will show how many connections are present on the machine, RDP, console etc. If there are multiple connections then this could be whats causing the problem. Hope this helps