After restarting primary hub, robot inactive alarm occurs for robot(s) already in maintenance schedule.

This is how maintenance mode works and why the problem is happening.

- maintenance_mode probe provides the maintenance info (device list, time and so on) to NAS probe.
- maintenance_mode probe uses UIM DB (MAINTENANCE_ table) to manage maintenance info. 

So NAS requires talk to maintenance_mode regular basis.
However, when the Primary HUB is restarted, NAS is started faster than maintenance_mode probe is started.
NAS attempts to talk to maintenance_mode probe but the talk is not established due maintenance_mode yet started.
NAS wait for a while (few minutes) for next talk retry.
Talk in between NAS and maintenance_mode established.
The maintenance info is not synced with NAS until talk is re-established in between NAS and maintenance_mode probe. 

Release: UIM 8.x or later

The workaround is to add the following line under <alarm_enrichment> section in controller.cfg in primary hub instead of disable/enable NAS manually. 

start_after = maintenance_mode