Orchestrator 2018.3 (actually upgrading to 2019.10).
Sometimes, when a job run immediatly after another one finishes, we got the Error “Executor start process failed, A specified logon session does not exist. It may already have been terminated”. Looks like the first RDP session did not manage to logoff correctly before the second job start.
We would try the UIPATH_SESSION_TIMEOUT increase fix but we can’t find on which machine to put this : Orchestrator server or UIRobot machine ?