Since v6.5 upgrade "Host hardware sensor status changed"

[Vitaliy S.]

Guys, the patch addressing host hardware sensors issue is already available, see this topic for more details: Patch #1 (build 6.5.0.686) [December 18, 2012]

[reizarf27]

Patch 1 still did not resolve this issue for me. All my ESXi hosts are still generating alarms in Monitor for an unknown host hardware status. For now, I have disabled the unknown state. Should I open a support request?

[Vitaliy S.]

This patch suppresses alarms triggering for stateless sensors only. Can you please take a look at vSphere Client to see what state you have there for the sensors in question?

[reizarf27]

In Monitor, the hardware alarm I am getting is for 'Proc 1 Level-1 Cache' being in an unknown state. In vSphere, the 'Read Policy' state of this hardware is currently showing as 'unknown'. I am getting this alarm for all the Processor cache hardware on all my esxi servers. I take it unknown is different than stateless?

[Vitaliy S.]

Yes, you need to investigate why these sensors are reported as "unknown" in vSphere Client and in Monitor Client. What is your ESXi server make and model? Is it a branded hardware?

[reizarf27]

They are all HP Proliant DL380's.

[Vitaliy S.]

Hmm...seems like you're not the only one who has HP Proliant with unknown sensors (check this thread for more examples). I wonder what other HP users have found to resolve this issue, probably they will be able to chime in to this discussion.

[ijefferson]

Happy New Year!

Just to say I've applied the patch today and all the stateless sensors have been removed except the CPU cache which is showing as unknown in Veeam One, but vCenter doesn't have anything listed next to it like the rest of them which the patch removed.

The only unknown listed there is like reizarf27 with the read policy.

Screenshot below:

[Vitaliy S.]

Hi Ian, thanks for the screenshot, I will forward this info to our R&D for further research.

[KiwiJJ]

Hi,
I have exactly the same problem as Ian. On all of my Dell R620 servers.

[Vitaliy S.]

John, can you please contact our support team and let them have a look at this issue a bit closer?

[KiwiJJ]

Hi Vitaliy,
I have raised a case, number 00174635.

thanks,

John

[GarethWilson]

i am also seeing these error being generated on my HP BL460 G6's & G7 's

These error only started appearing after installing patch 686 yesterday

[Vitaliy S.]

Can you please let our support team take a look at this problem live?

[GarethWilson]

Yes No problem i am more than happy for them to remote in and have a look at this issue if they want

Cheers

Gareth

[GarethWilson]

Did you want to get them to contact me (we do have a support contract for veeam B & R and veeam one) on 01792 602020 ?

[Vitaliy S.]

You need to open a support case on this issue and give a reference to this thread, and I will ask to escalate it to the dev team for further investigation.

[GarethWilson]

Case number 00175512

[DaveCox]

Is there any update on this? I am running 6.5 with update 1; DL380 G5; vcenter is reporting "unknown"; veeam is reporting "unknown".

[Vitaliy S.]

Well... If vCenter server reports unknown status there is little we can do. You should check it with the hardware vendor technical team or disable this alarm for the sensor in question.

[GarethWilson]

yes i have currently got a call open with vmware but i am not holding my hopes up.....Surely as a vmware partner veeam could be talking to vmware about this issue ?

[Vitaliy S.]

You're right, we always try to work with VMware on the issues we can resolve together, but keep mind that we do not control the way how VMware retrieves this information from the ESX(I) hosts. I believe it shouldn't be hard for them to nail down this issue. Also be aware that the list of hardware sensors solely depends on your hardware make and model, so it might also be a firmware issue.

[DaveCox]

My confusion then. From your description (below) i thought that if vcenter was resporting unknown (screenshot from Ian Jefferson) this is "stateless" and your hotfix would sort it. That didnt make sense to me.

1. Hardware sensor alarm is triggered for "stateless" sensors, the sensors that do not have any state in vSphere Client (see screenshot from Ian Jefferson, vmnic sensors).
2. Some sensors might have "unknown" state in the status column in Monitor Client while having status "Normal" in vSphere Client (and vice versa).

Our R&D team has found the way to resolve the first issue with "stateless" sensors and is already working on the hotfix.

[GarethWilson]

The only fix seems to be to disable these alarms on the processor CPU cache...

[Vitaliy S.]

From your description (below) i thought that if vcenter was resporting unknown (screenshot from Ian Jefferson) this is "stateless" and your hotfix would sort it. That didnt make sense to me.

Not exactly, "unknown" is still a state for the hardware sensor. Stateless sensors are sensors that do not have any state at all.

[cstemaly]

What I did to get rid of the error was to edit the alarm settings, and then disable the "unknown" section (at the bottom) which causes the warning. This way, I figure that I would still be alerted if a power supply or CPU went out. Is that thinking correct?

[Vitaliy S.]

Yes, that's correct. Also you can disable this alarm for a particular hardware sensor if you navigate to Hosts->Hardware tab within the Monitor Client.

[stevenrodenburg1]

The only fix seems to be to disable these alarms on the processor CPU cache...

Yep. I've seen this at several customers (using several different HW vendors/models) after upgrading to Patch1.
Disabling the CPU cache alarm always reverts those servers back to "green". It is a pain in the ass however if one has to do it for each and every server. I have not found a way to just disable that CPU cache checking in Alarm Management -> VMware -> Host -> Host Hardware Status.

[Robert M.]

Hi Guys,
is the last entry of Stevenrodenburg1 the "final statement" ? -> just for Information -> I checked the "unknown Status" on an HP BL460C Gen8 on which is running ESXi 5.1.0-799733 and the same was happening. In this post section, I found that there some cases were opened, but without success for a fine solution - correct ? (Only the hint for disabling the "Alarm or Hardware Sensor-Settings") - but maybe someone got a new solution ?

[Vitaliy S.]

Robert, there is no "solution" at the moment as we retrieve status for the hardware sensors from VMware. If status is reported as unknown, then it also gets reflected in the UI.