VMware - persistent "Host memory status" alarm in vSphere

Question

I have a particular VMWare ESX 4.1 host that has a very persistent "Host memory status" alarm.

This is running on an HP ProLiant DL360 G7 server. The HP ILO and System Management agents don't know any errors. If I clear the alarm in the vSphere client, it returns within a day. I've tried reseating DIMMs, however, the error does not indicate a problem with a specific module.

There's another host in the cluster with an identical configuration. It's not exhibiting any issues.

Any thoughts?

This is touched on briefly on other forums (and here) with no clear resolution.

Solution 1

whens the last time you've updated the HP ESXi stuff? There could be a bug in thier HW pack that's been resolved, or provides more info.

Solution 2

I experienced the same issue and this article solved it. I ended up having to perform all 9 steps.

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2011531

To resolve this issue, you must clear warnings and errors from the Hardware Status tab in VMware vCenter Server.

To clear warnings and errors from the Hardware Status tab:

1. Go to the Hardware Status tab.
2. Click the System event log view.
3. Click Reset event log.
4. Click Update. The error clears.
5. Click the Alerts and warnings view.
6. Click Reset sensors.
7. Click Update. The memory clears.

Note: If the error does not clear, connect, connect to the host via SSH. For more information, see Enabling root SSH login on an ESX host (8375637).

8. Run this command to restart the sfcbd service:
   services.sh restart

   Run this command to restart the service in ESX:
   /etc/init.d/sfcbd-watchdog restart

9. Click Update. The error clears.

Solution 3

I faced with the same problem recently on my HP Proliant DL360 G9 and VCenter 5.5. Here is the magic sequence from Cannot clear Hardware Status warnings and errors in VMware vCenter Server 4.x and 5.x that helped:

To clear warnings and errors from the Hardware Status tab:

1. Click the Hardware Status tab.
2. Click the System event log view.
3. Click Reset event log.
4. Click Update to clear the error.
5. Click the Alerts and warnings view.
6. Click Reset sensors.
7. Click Update to clear the memory.

Note: If the error does not clear, connect to the host via SSH

8. Run this command to restart the sfcbd service:

   services.sh restart

   Run this command to restart the service in ESX:

   /etc/init.d/sfcbd-watchdog restart

9. Click Update.

See also Warnings in the Hardware Status tab of the ESXi 5.x host fails to clear documents for more details.

Comments

[Admin]

Anything in the IML logs? both on the same firmware? sounds like a HW issue to me

[Admin]

IML is clean. SMH is clean. No LEDs/indicators on the front display.

[Admin]

I could swap DIMMs or order warranty replacements. But since the error doesn't specify a slot...

[ewwhite]

Updated to the newest ESX 4.1 update Wednesday. I'll try updating the PSP now.

[ewwhite]

It's been a day, and I've had no errors since. So updating the HP ProLiant Support Pack seems to have helped.

[slm]

Welcome to Server Fault! Whilst this may theoretically answer the question, it would be preferable to include the essential parts of the answer here, and provide the link for reference.

[Stone]

Reset event log did the trick for me! I was scratching my head over this discrepancy for a day now, but finally it it solved! Thank you!