Chrome - process failing...sometimes?

[MattMunch1]

Hey guys, When spying elements in chrome, it seems that sometimes the elements will be picked up and work as intended, and other times it's like browser mode is completely disabled. Nothing is possible to be spied when using Application Modeller, regardless of website. I'll run the process in process studio multiple times, and the first few times it'll work, and then it'll fail out of no where. Literally running it, resetting it, and running it again - 20 times or so. Some fail, some work. Increased wait times don't affect the success rate. We tried 6.4 with Chrome version 71, and that didn't work. Tried the new 6.4.2 with Chrome version 71 and 72, neither of those worked. Tried disabling network connectivity in chrome://flags on both chrome 71 and 72, that didn't work. And to top it off - it now looks like the BP Extension is gone from our browsers, and it's not available on the webstore. Reinstalling 6.4.2 did not bring it back. Does anyone have any idea what's going on with Chrome?

[krishnavenijala]

same with me as well,

[Christer__Tellefsen]

You must have Blue Prism 6.4.2       http://portal.blueprism.com/customer-support/support-center#/path/1319712802   Why is my Process no longer able to spy or interact with Chrome web browser? Share this answer The way Chrome integration currently works is, after a fresh reboot, once a Process is run that interacts with Chrome, a connection is made with Chrome over port 31924. Even after that Process has completed, the connection to port 31924 will still be open and is not disposed of. When performing testing of a Process via Studio, the Process is run via the local (private) Resource. As described above, the local Resource would initially open the connection, and that local Resource would then hold on to that connection. If you subsequently attempt to run the Process from Control Room to a Runtime Resource running locally on the same machine, as the connection had already been opened by the local Resource, a new connection cannot then be initiated, which causes a failure. For this reason, only performing a reboot of the Resource machine will free the connection and allow the Process to run again via either Control Room or Studio. Our expectations are that you should only encounter this issue during the testing phase of the Process. The reason for this is because you will be switching between running the Process within Studio and Control Room. We don’t expect the issue to present itself within a Production environment as the Process would only ever be run from a Runtime Resource. To prevent any further issues going forward, it is recommended that a separate Resource is used to perform any Control Room testing. This will ensure that both Studio and Control Room have their own connection for the Chrome integration.