Over in the forums, goolj10 asks:

"It seems to me that if your management server goes down, then there is no way to a) start new VS's, b) restart a VS that failed either due to an internal error or because the hardware it was running on has failed. Is this true? If so, what recommendations do people have for overcoming this single point of failure?"

cbarclay responds that Virtual Iron recommends running an Active/Passive cluster to protect against these types of issues. He continues with:

"Follow the installation and administration guide to install two management servers. Make sure both management servers have identical hardware and are configured to boot from the same LUN. We recommend that the LUN is mirrored using hardware RAID.

A network power controller and network monitoring software can fully automate Virtualization Manager failover. Network monitoring software can monitor port 80 or 443 to determine if the Virtualization Manager is healthy. Alternatively, a heartbeat script can be written using the command /etc/init.d/vivmgr status.

When the Virtualization Manager experiences a hardware or software failure, either automatically using a network power controller, or manually, the active server can be powered off and the passive server powered on. Once this server boots, the Virtualization Manager will be back online without impacting the availability of any virtual server."

Our forums contain a bunch of tidbits like this and the community is always happy to answer any questions you have.