HP Virtual Connect Manager Version 1.24 Release Notes and Compatibility Matrix

Known issues 12
any disruption to the servers, but the administrator should ensure all servers have the correct MAC
addresses and WWNs. Powering off the server will clear the "profile recovered" state. If any servers
are rebooted or power cycled while the credential recovery is occurring, the MAC addresses and
WWNs may be returned to the factory default settings.
• If an enclosure import is attempted with a server blade in a failed state, the Virtual Connect Manager
might report an erroneous error. If the import completes, close the browser and log in again to verify
the import was successful. The working state of all server blades can be verified using the Onboard
Administrator.
• Changes in the server ROM-based setup utilities (RBSU) can override Virtual Connect server profile
settings for PXE (enabled or disabled). If a change is made in RBSU, that change is not reflected in
Virtual Connect Manager. If PXE is set through RBSU, HP recommends using the "Use BIOS" setting.
• RDP "rip and replace" does not work properly when using VC-administered MAC addresses
because RDP is tracking the server blades by MAC address rather than by the primary lookup key
(which can be set to serial number, MAC address, UID, or Asset Tag). To work around this issue,
manually redeploy the server blade.
• For VC-Enet modules, do not use the Get-Bulk feature with SNMPv2c MIB walk. The MIB walk
operation halts prematurely before running through the entire MIB. When using SNMPv2c protocol
for MIB walk, make sure that the Get-Bulk option is not enabled.
• When scripting CLI commands, only a single scripting client should perform remote management
operations to a remote Virtual Connect Manager. If multiple scripting clients are used to perform a
heavy load of CLI commands to a single Virtual Connect Manager, some management commands
might fail, and in some cases, the primary module might need to be reset to properly recover.
• An unsupported state occurs when a VC Ethernet module is replaced physically with a non-VC
Ethernet module without administrative removal of the module from the VC domain. During this state,
if either the Help/About HP Virtual Connect Manager page or Firmware Management page of the
VCM GUI is opened, the string, "Loading...," appears indefinitely instead of the firmware version
number for each module in the VC domain.
Fibre Channel and SAN boot
• If updating from Virtual Connect Manager v1.10 firmware, the VC-Enet module firmware must be
updated before the VC-FC module firmware. Otherwise, the VC-FC module firmware might not
appear to be successfully updated.
• To properly boot from SAN when using Linux and VMware ESX 3.0.1, change the QLogic
QMH2462 4Gb FC HBA connection option to "point-to-point only" in the QLogic BIOS configuration
utility. The Emulex LPe 1105-HP 4Gb FC HBA does not require using the "point-to-point" connection
option to properly boot from SAN.
• Virtual Connect Manager does not provide an interface to force 1-Gb speed on the VC-FC module
uplink and downlink ports. However, the ports will auto-negotiate properly to 1 Gb.
• The Fibre Channel login sequence can be delayed by approximately 20 seconds when more than
one server blade HBA is aggregated through the same VC-FC uplink port attached to a Cisco switch
and these HBAs are in the same zone. To avoid this problem, do not place server HBAs in the same
zone on the fabric switch.
• When configuring Trap Destination IP addresses for the VC-FC modules, be sure that trap
destinations are capable of supporting SNMPv2 type traps.