I now conducted an update on another remote device and observed the same problems, now well documented, I hope:
Device C: Physical machine: W10-Laptop at private Home, CB-RA-2.1 at that time, sitting on the same Desk with machine B so I can execute remote stuff and observing everything in parallel.
Chain of events:
Establishing connection from B (local) to C (remote), B already on RA-2.2, C still on RA-2.1. Using known credentials, 9-digit-ID ("4xx-xxx-xx0") and keyfile.
Connection established and works.
Right after that, via remote ≡ menu check for updates and acknowledging the expected connection abort:
After that, the connection closed:
Now having a look at the Laptop's screen (device C) directly. The previouslly mentioned error dialogue almost immediately showed up, now here in full detail screenshotted:
("The process can't access the file c:\…, bc. it's in use by another process")
Funny enough, the update runs in the background and after clicking away the error, RA restarts automatically – and now with a different ID ("8xx-xxx-xx2", not the previous "4xx-xxx-xx0"), and without encryption and unattended access enabled:
Clicking encryption, generating a key and trying to save it yields the following error message:
("Path not available. (X) "C:\windows…" is not available. … yadda yadda yadda
Confirming that error with OK and after that shutting down RA and restarting it solves the problem. It is then possible to generate a key, save it, transfer it to the local machine B, entering the new ID and importing the key and then establishing a connection and working via it.
So there seems to be an issue with the update process started remotely via a RA session on the remote machine. I'll try out the outcome when updating RA locally on a machine D first and then connecting there from, say, machine B.
Hope this helps better.