If I have a Orchestrator monitoring ‘X’ robots, and the Orchestrator falls down (the server is down), how the robots run (or not run)? The robots stop inmediatly? they run to finish the process and after they send a report/logs to the Orchestrator (when this will be up)?
Robots will still be running, at least in 2016.2. Don’t know about 2017.1.
Logs will be stored only locally and all queue and asset related activities will fail (obviously). This also means that currently processed transactions will be left in InProgress state.
But you will have a chance to at least store whatever was happening locally and possibly do some compensation on the incomplete transaction.
It the same behavior described by Andrzej for 2017.1 version too. Logs will be stored in execution logs and also cached(on local machine) so when the server is up again the service will send the logs to Orchestrator.