HP Matrix Operating Environment Software - Manual (Page 83)

Table 5-1 Troubleshooting an inoperable server
Suggested action Possible cause
Perform one of the following actions:
• Reverse the operation which made the underlying VM
host or blade unavailable for management operations.
If the underlying resource is a blade, consider restoring
the domain group or bringing the domain group back
from maintenance mode. For a VM Host, reregister the
VM Host with Insight Control virtual machine
management, or restart the VM Host.
• Unmanage the inoperable logical server.
The VM host or blade resources associated with a logical
server have been made unavailable for further
management operations. Examples are:
• Using Virtual Connect Enterprise Manager to remove
a domain group from management or to move a
domain into maintenance mode
• Using Insight Control virtual machine management
to unregister or shutdown a VM host
Perform the following sequence of steps:
1. Unmanage the inoperable logical server
2. Determine which of the underlying resources is the
intended instance, and delete the other instance. Delete
the underlying VM with Insight Control virtual machine
management, or delete the blade instance with using
VCEM.
3. Verify the integrity of the remaining instance; if
necessary, restore the remaining instance from backup
and unmanage and re-import the recovered instance.
If the integrity of the remaining instance cannot be
verified then remove it and recreate.
Both the blade and VM guest resources of a fluid
cross-technology (portable) logical server have been
activated and are operating simultaneously. When blade
or VM guest resources associated with logical servers
are activated using VCEM or Insight Control virtual
machine management, dialog boxes appear warning that
such activation is not recommended.
Unmanage and re-import the impacted logical server. Inoperable logical server appears immediately after a
failure of the LSM service or of the entire CMS. Partially
completed operations, such as move operations, will not
resume after the LSM service or CMS restarts, and the
CMS may not be able to synchronize the state of the
recovered underlying resource with the logical server.
VCEM provides a warning when attempting to perform management operations on a blade
associated with a logical server. Performing such operations within VCEM may lead to inoperable
logical servers.
The warning message reads as follows:
IMPORTANT: VCEM has detected you may be using other products such as
HP Insight Dynamics-VSE. Assigning this Server Profile can make
it inconsistent with the upper level manager. It is recommended
to use the upper level manager to assign this Server Profile.
Type 'YES' to proceed with this operation.
Correcting problems powering on a logical server
Occasionally a Virtual Connect logical server or virtual machine logical server fails to power on
and returns an error. If you encounter this problem, check the firmware versions and hardware
as described below.
Suggested Action: Check firmware versions Check the recommended firmware versions of
Virtual Connect Manager, Onboard Administrator, Integrated Lights Out (iLO) and server blade
BIOS. A firmware version different than the recommended version could cause a power request
to fail.
Suggested Action: Check hardware using Onboard Administrator Most power issues are related
to the hardware itself. For example, power problems can be caused by temperature, fan speed,
fans not installed in an effective configuration, insufficient power because of power supply
location, Blades and Interconnect Modules health problems, or device failure errors in general.
These problems can be identified using the Onboard Administrator Rack Overview page (available
in HP SIM through Integrated Manager) or Onboard Administrator Insight Display.
Correcting problems powering on a logical server 83