HP Systems Insight Manager - Manual (Page 157)

Solution: Remove all systems associated with the MSCS cluster (this includes the cluster alias, the
cluster nodes, and any VMs that might be running on the cluster), and then re-discover the cluster
(and any VMs) without discovering the service. For example, add the IP address of the cluster
service to the ping exclusion range on the general discovery settings page.
-------------------------------------------------------------
When the number of objects to be collected on any given storage array exceeds the value of
~2500, it is possible that data collection will fail for the array.
Solution: This is due to a default timeout value of three hours for this operation. There are two ways
to resolve the issue:
•
Configure the array to have a maximum of 2500 collectible objects (volumes, disks, ports,
and so on).
•
Increase the timeout value, depending on the configuration, to an optimal value by modifying
the configurable parameter Storage_DC_Timeout which is set to a default value of 10800
(equals 3 hours), with the mxglobalsettings command.
◦ Example to retrieve the current value:
mxglobalsettings -ld Storage_DC_Timeout
◦ Example to modify the value to 4 hours (14,400 seconds):
mxglobalsettings -s Storage_DC_Timeout=14400
It has been observed that data collection can take 3 hours and 37 minutes on an XP24000 array
that has ~3200 lives, 40n network ports, and ~150 disks.
-------------------------------------------------------------
If an HP Logical Server that is created in Insight Dynamics is given the same name as the operating
system host name of the blade on which it is applied, then the logical server is deleted when the
blade is rediscovered.
Solution: To avoid this, be sure the logical server name is different from the host name of the blade.
-------------------------------------------------------------
If you:
•
Configure an HP Serviceguard package as an HP virtual machine which is hosted by two
different HP virtual machine hosts,
AND
•
Complete a failover of the package from one HP virtual machine host to the other one,
AND
•
Re-identify the package in Systems Insight Manager,
you might find that the package is not associated with the correct HP virtual machine host.
Solution After a Serviceguard package fails over from one host to a different host, you must
re-identify the HP virtual machine host system in Systems Insight Manager to see the correct
association.
-------------------------------------------------------------
Orphan systems are appearing after I run discovery.
Solution: To prevent orphans from appearing in the future, review events and remove the system
that you have migrated to a new system type before rediscovering the new system after it is booted.
This is true of systems moving in or out of a virtual system environment.
-------------------------------------------------------------
If you see the preferred system name is different than the host name of the full DNS name for an
HP virtual machine guest, it is because the preferred system name is specified in the virtual machine
package.
Discovery 157