Monday, June 30, 2014

Dell MP 5.1: Where Are The Dell OEM Servers?!

Issue
For a customer the Dell MP (only the Scalable MP, not the Detailed MP…) is in place. However, many Dell servers aren’t discovered, even though all prereqs are in place AND operational. But no matter what, these servers weren’t do be discovered as Dell servers. Time for some troubleshooting.

Cause
The Dell Server MP (Dell Windows Server (Scalable Edition) contains some Discoveries. The Discovery Dell Server Discovery is the main Discovery here, targeted against the Windows Computer class. This Discovery runs once per 24 hours.

This Discovery can be modified through the regular mechanism of setting Overrides. In this case I set the Discovery to run once per minute for a particular Dell server having these issues AND I enabled logging of this Discovery by setting the Parameter Name Log Level to 1:
image

Because of this the Discovery will create the log file ‘C:\Windows\Temp\Discovery_DellServer.log’. Soon when the Overrides were distributed to this server and processed by the Microsoft Monitoring Agent, the log file was created and neatly filled with information. And this is what came out of it:
image

So besides the hardware manufacturer Dell, there is also a manufacturer DELLOESad smile.

Time for a deeper investigation of the Discovery involved. I exported the MP to a plain XML file and used Notepad++ to open it. Soon I found the Discovery and it’s script, DellServerDiscovery.vbs. This is a huge script containing many functions, among them the function GetManufacturerName:
image

So this function runs this  WMI query: Set wbemObjectSet = GetObjectSet(wbemServices, "select * from Win32_ComputerSystem", "Manufacturer", "Win32_ComputerSystem"). Since the log file contains the entry INFO: HardwareManufacturerName Name retrieved successfully, I know this function run just fine. However, the value it contains isn’t accepted… Back to the Discovery script.

And yes, here I found the pain: Instr is case sensitive:
image

So this causes the Discovery not to work for servers with DELLOE as hardware manufacturer…

Where does DELLOE come from?
As is turns out, DELLOE is related to Dell servers running an OEM BIOS. Apparently this information is piped into WMI as DELLOE.

However, installing a Dell BIOS on servers like these isn’t the way to go since I don’t have any idea to what extend it will impact the operations of this server in a negative kind of way.

Solution?
As this moment I am thinking about multiple ‘solutions’ or better, workarounds. Ranging from dirty to acceptable but involving a lot of work. First I have to discuss them with other people and look into them. When I know more, I’ll update this posting accordingly. For now these Dell servers running an OEM BIOS won’t be covered by SCOM Sad smile.

When Googling (sorry, here in Europe I don’t use Bing) with this query Hardware Manufacturer Name query : DELLOE, I found that I am not the only one having this issue:
image

Recap
Looking back at previous encounters I had with the Dell MP I must say I am not surprised. But ‘only’ disappointed…

The Dell MP has become far more better compared to the past but it’s still a MP to be careful with and not to load all available Dell MPs since before you know it, you’re in deep troubles.

1 comment:

PcD said...

He Marnix,

version 6.2 of the management pack solves the problem.

Cheers