We've had an issue where the appliances had ran out of disk space, and AA became unstable. On another occasion, it filled up for another reason and the portal fell over. For any over VM, we would add them to our monitoring so we could get an early warning, but that's currently not an option. Given how essential this service now is (and this happened most recently during exam resits), it's essential we either the university or AppsAnywhere actively monitors the appliances in order to be more proactive. For example: SNMP can be configured without requiring AA to monitor it on site AA uses their own system for monitoring the appliances The appliance sends alerts to AA and/or the university (although that can have its own flaws) More comprehensive monitoring is available from the Admin portal (although that can have its own flaws)