Hardware Health Observations

Deltona

Hi,

Ever since SAM 5.5, I've noticed that when adding a Node to Orion and listing resources, the Hardware Health monitoring option is present.

This wouldn't seem strange and yet it does, considering the Node being added is not an HP, Dell or IBM server. The Nodes being added are indeed physical hardware of alternative brands.

However, after the first scheduled poll (circa 9 minutes), the Hardware Health resources disappear. When listing resources again, the option to monitor Hardware Health is now missing.

During the short period between adding the Node and polling the node, the Current Hardware Health resource displays relevant component information such as Fans, Power Supplies, Disks etc.

I've tried understand this behavior but only managed to confirm that if the Nodes have a CIM service running on them then Hardware Health monitoring would be picked up.

This is the case for systems with and without "agents" installed. Our BSD systems for example have CIM service installed and running as part of the OS. As with Intel servers, if they have the SNMP Management CIM installed, they'll get picked up by Hardware Health as well.

Now I'd really like to know why does SAM automatically disable Hardware Health monitoring of these systems, even though it works.

This goes without saying, but I'd REALLY like to monitor HH on these "non HP, Dell, IBM" systems. Is there any way I can force SAM to "keep" monitoring Hardware Health on these systems?

I take it only systems where Vendor is equal to one of these three is accepted. If so, why the favoritism?

@aLTeReGo ?

aLTeReGo

There are several different implementations of CIM. The CIM standard simply defines how data should be structured, not how it is accessed. For example, WMI is Microsoft's implementation of CIM, which is based on DCOM. VMware on the other hand has their own CIM implementation based on HTTP/XML. I'm not sure what manufacturer the server is, but if you're willing to provide diagnostics immediately following the node being added and collecting Hardware Health information, we would likely be able to determine what is triggering this from being detected in the first place, and what's causing it to disappear.

It is not our intent to block systems from working with SAM's Hardware Health monitoring, but we need to validate and verify each vendors implementation individually before we can support it in any official capacity. The current checks implemented in SAM are designed to ensure that the Hardware Health option only appears for nodes types that are officially tested by our internal development and QA team, and supported by our engineers.