Servers: Windows 2008 R2, ZCM 11.2.1, Agents Windows 7, ZCM 11.2.1, DB = Oracle 11g. Active Directory environment
Does anyone know of methods to improve the ZCM client PC login time when the backend DB is unavailable?

We have had some service outage events recently which have resulted in agents having long login delays during the outage events.
We use passive login and so expect at a windows login to have Zenworks automatically login using the windows credentials.
We've improved the login experience when on unknown networks and if 1 or more ZCM servers are down but have been unable to improve the situation if the ZCM servers are up but the oracle database is down.

When the backend database is down the agents do not behave in the same way as for a 'bad' front end ZCM server.
For "unknown network" (and "all ZCM servers off") we have the login experience down to less than 30 seconds the login when the DB is down is still more than 5 minutes - without an opportunity to cancel.

This is giving Zenworks some very bad press with our users.
Any possibility to add a cancel button to the ZCM agent login after a period of no success?

Please don't tell me that the Database should simply always be up! We live in complex environments and must plan for outcomes during ANY service outage. There is no need for windows login to be stopped just because a part of the ZCM environment is down.

Settings that helped with the login times for front-end outages (or unknown networks)
1. Set each location to have only 2 ZCM servers, unknown loadtions to have none
2. Set retry options (ZCM: Configuration | Device Management | Zenworks Agent | General)
- times to retry requests to a busy server: 2
- Initial retry request wait: 1
- maximum retry request wait: 5