ESXI + SNMP

[venom1318]

I am trying to setup SNMP monitoring on an esxi hosted system. The reason vsphere alone isn't enough is because this is for one of a couple hundred clients, and we would rather the SNMP monitoring be tied into our central monitoring system, not simply receive emails from vsphere when a problem occurs. So once I track down the correct OID, everything else is in place to monitor this automatically for status changes.

I have gotten to the point where I can run an snmp walk against the esxi host and receive information, but so far I have been unable to retrieve hardware health information. At a minimum I would like to monitor the array health. I would think that somewhere in the esxi snmp information that is returned the health status information that can be found in vsphere under the configuration tab would be available. But as of yet I have been unable to track this down.

I am using the 5.5 version of the vmware mib files. Not sure if anyone can point me to a certain MIB that should help me find this data.

Hopefully that was enough information. I really think I'm just not hitting the right MIB file to read the data out from the SNMP walk.

[BenLiebowitz]

Have you gone though and enabled SNMP on the hosts and configured the trap community, etc? 

VMware KB: Configuring SNMP Traps in ESXi/ESX

[venom1318]

Yes. All of that has been done.

As I said, I can perform an SNMP walk and receive a response and with some of the MIBs loaded I can see information about the virtual machines.

What I want/need to see is the same information displayed in vSphere under the host heath status.

I would have thought that since all of that information is available in vsphere, one of the MIB files should contain the information to parse that out of the SNMP walk results. I just haven't been able to find it.

[jrmunday]

Have you queried these MIB files;

VMware MIB Files

Perhaps you need the vendor specific management bundles installed to get additional MIB's installed - for example, Dell Management Bundle or HP Offline Bundle?

[venom1318]

I looked through the documentation for the MIBs. It doesn't really sound like any of them contain the information I need.

I've installed another SNMP tool that allows me to load as many MIBs as I need. I've loaded all of the HP and the VMware MIBs and I still cant find anything in my SNMP walk about the physical hardware health.

It is ridiculous that this isn't available when the information is right in vSphere. There has to be some way to get at the information if vSphere can see it. I'm just running out of ideas.

[venom1318]

Even if I could find a general alert that vSphere detected a hardware problem of any kind, that would be enough to at least trigger us to check on the server.

[MRCRDN]

Hello,

I'm really interested in this question too.

I have multiple ESXi host with multiple VMs.. but I mainly want to monitor the Health Status of the Host.  (Power Supply (Status and Wattage), FANs, Temperature, disk array status).

I'm using CACTI monitoring to graph all this, all I need is it correct OID.

Thanks!

[MRCRDN]

Ah!.. I think I found some values... not exactly what I need but close.

iso.3.6.1.2.1.25.3.2.1.3.16 = STRING: "POWER Power Supply 1"

iso.3.6.1.2.1.25.3.2.1.3.17 = STRING: "FAN Fan Block 1"

iso.3.6.1.2.1.25.3.2.1.3.18 = STRING: "FAN Fan Block 3"

iso.3.6.1.2.1.25.3.2.1.3.19 = STRING: "FAN Fan Block 4"

iso.3.6.1.2.1.25.3.2.1.3.20 = STRING: "MEMORY Memory"

iso.3.6.1.2.1.25.3.2.1.5.16 = INTEGER: 2

iso.3.6.1.2.1.25.3.2.1.5.17 = INTEGER: 2

iso.3.6.1.2.1.25.3.2.1.5.18 = INTEGER: 2

iso.3.6.1.2.1.25.3.2.1.5.19 = INTEGER: 2

iso.3.6.1.2.1.25.3.2.1.5.20 = INTEGER: 2

        SYNTAX   INTEGER  

             unknown( 1 ) ,

             running( 2 ) ,

             warning( 3 ) ,

             testing( 4 ) ,

             down( 5 )

         DESCRIPTION

              "The current operational state of the device

              described by this row of the table.  A value

              unknown1 indicates that the current state of the

              device is unknown.  running2 indicates that the

              device is up and running and that no unusual error

              conditions are known.  The warning3 state

              indicates that agent has been informed of an

              unusual error condition by the operational software

              e.g. a disk device driver but that the device is

              still 'operational'.  An example would be high

              number of soft errors on a disk.  A value of

              testing4 indicates that the device is not

              available for use because it is in the testing

              state.  The state of down5 is used only when the

              agent has been informed that the device is not

              available for any use."