register
Contributor
Contributor

vmware-hostd high cpu utilization

esx 3.5 update 3.

I received alerts from a couple of hosts in my DR cluster today for high cpu utilization. The problem occurs even if no VMs are running on the host. The other host in the cluster is fine. There is one vmware-hostd process running consuming 50 - 60% of the cpu. I have tried restarting mgmt-vmare, vmware-vxpa & even rebooting one of the hosts. However the problem keeps returning. Any suggestions?

thanks

0 Kudos
2 Replies
svaradha
Enthusiast
Enthusiast

Hi. I too have noticed this rather often after updating to U3. This affects all hosts now and then...

Regards,

-Satheesh-

If you found this information useful, please consider awarding points for "Correct" or "Helpful" answers/replies. :) Regards, -Satheesh-
0 Kudos
register
Contributor
Contributor

I found the issue (bug?). I checked the hostd.log files. I had updated some resource limits earlier in the day. We replicate VMs to our DR site, so the replicated NFS mounts containing the VMs are Read Only. The task to modify resources shows sucessful, but the hostd.log was logging numerous errors every second because it was unable to update. I had to make the NFS mounts temporarily Read Write, then restart the mgmt-vmware daemon. Otherwise, it would log this forever:

GetPropertyProvider failed for haTask-ha-root-pool-vim.ResourcePool.updateConfig-539348289

If you try to update the Resources of a VM on a Read Only volume, shouldn't the original update either fail immediately, or at least timeout & fail?

0 Kudos