Monitoring Local Disk on ESXi (DL380 G5) on HP SIM 5.2

[Tom Parker_1]

Just a clarification on the last post. It is update 4 that I just installed today.

[Erik Persson]

Nice to see that ESXi finally can monitor the health of the Smararrays but now I have kind of a follow-up problem.

We are using Nagios do do our monitoring and I am not that familiar with WBEM. I found a check_esx_wbem.py which uses the pywbem Python module.

The check script works insofar that it will report power supply failures and such but it will not report problems with the disk controller even though it is clear that ESXi does notice if I for example yank a disk (I get critical alerts in the VI client).

Analog to SNMP my suspicion is that the check script does not ask for the proper WBEM classes as far as the smartarray is concernered. Anyone here who has an idea what the correct class may be called?

This is the code block in check_esx_wbem.py that seems relevant:

# define classes to check 'OperationStatus' instance
ClassesToCheck = [
'CIM_ComputerSystem',
'CIM_NumericSensor',
'CIM_Memory',
'CIM_LogicalDevice',
'CIM_Processor',
'CIM_RecordLog',
'OMC_DiscreteSensor',
'VMware_StorageExtent',
'VMware_Controller',
'VMware_StorageVolume',
'VMware_Battery',
'VMware_SASSATAPort'
]


Note: I guessed on "CIM_LogicalDevice" and added it myself but that only caused the check script to report the second NIC to be unplugged (correctly).

Can I query the WBEM interface for relevant classes in some way? I am not a Python programmer?

Thanks,

/Erik

[Bahti]

Hallo Eric,
we have the same ideer and the same Problem,
I testet vollowing CIM´s
ClassesToCheck = [
'CIM_ComputerSystem',
'CIM_NumericSensor',
'CIM_Memory',
'CIM_Processor',
'CIM_RecordLog',
'CIM_LogicalDevice',
'CIM_StorageExtent',
'CIM_Battery',
'CIM_Controller',
'CIM_CoolingDevice',
'CIM_DiskDrive',
'CIM_Fan',
'CIM_MultiStateSensor',
'CIM_PhysicalComponent',
'CIM_PowerSupply',
'CIM_Processor',
'CIM_Sensor',
'CIM_Slot',
'CIM_StorageVolume',
'CIM_SASSATAPort',
'CIM_AllocatedFromStoragePool',
'CIM_HostedStoragePool',
'CIM_StoragePool',
'OMC_DiscreteSensor',
'VMware_StorageExtent',
'VMware_Controller',
'VMware_StorageVolume',
'VMware_Battery',
'VMware_SASSATAPort',
'VMware_HostedStoragePool',
'VMware_StoragePool'
]
but not one will detect a RAID error !!
has someone hir an idea or solution ???
webm or possibly also on ILO2 or ....
HP Proliant DL360 G5
Smart Array p400i
ESXi 3.5 U4 HP-image

[Goran Karlsson]

Have you tried the new HP release of ESXi?
"ESX 3i Update 4 with HP Managements Agents"?

[Bahti]

Yes I dit testet with
ESXi 3.5 U4 Hp-Image Build 153875
and I can see to test the Warning Message in VM-Ware Infrastrukture Client (see attached screenshot)

[Donutman]

I downloaded the "ESX 3i Update 4 with HP Managements Agents". Just tested it on a DL180G5 with a PCI-X Smart Array P600 controller and it shows up under 'Storage' in the ESXi health status fine, just like the poster above me shows for his P400 controller.

[ITRegister]

Can someone tell me if you can use the ESXi Update 4 ISO to upgrade an existing HP server or do you have rebuild it from scratch? Can't find any doco on this.

[Jools Chesters]

If you have an HP USB key you can write an updated image to it, instructions here:

http://blog.mechanised.com/2008/07/how-to-create-your-own-bootable-esxi.html

Personally I have also noticed on our BL460c's that they are reporting the E200i, attached is screenshot from a diskless server.

But we also have an ML350 G5 with the same controller that is not reporting anything?

[ITRegister]

To clairfy my last post, we already have ESXi HP Update 3 installed on a DL360 G5 with E400 controller, and want to upgrade it to ESXi HP Update 4.

Can you do that using the ISO from VMWare or does the server have to be rebuilt from scratch?

[Jools Chesters]

Just use VMware Infrastructure Update, it'll need a reboot afterwards though.

[Ligeret]

Hello,

The python script can monitor the RAID error as long as ESXi do.
I successfully test it on DELL PowerEdge servers but I don't have any Proliant server to perform tests.