We just upgraded our Cisco UCS environment to 2.0(1m) software. Since we upgraded we have had issues with the entire UCS rebooting each night. I just finished a TAC case with Cisco, they have not yet assigned this a bug ID but are now aware of what's happening. We've had the UCS loaded into the NCM even though NCM cannot backup the configurations it was doing inventories. The issue is IF NCM does an inventory on the UCS systems, whether manually or scheduled, it will trigger a memory problem in the UCS and they will both reboot. So, if you upgrade your UCS systems to 2.0(1m) you will want to remove them from NCM first. Cisco believes this may be linked to a previously patched bug where multiple simultaneous SNMP "gets" will trigger the memory leak.
