Errormessage about disabled VM autostartup & vSphere HA cluster

  • 27 May 2014
  • 3 replies
  • 1236 views

Badge +3
Hey guys,

We are busy building a Nutanix 3.5.3.1 infrastructure with vSphere 5.5. We have six nodes in two blocks, and the design consists of 2 Nutanix clusters with each its own vSphere cluster. HA is enabled on each cluster.

We are getting an error on both Nutanix clusters in PRISM UI:

Virtual Machine auto start is disabled on the hypervisor of Controller VM

Nutanix best practices recommend to enable "Start and Stop Virtual Machines with the system" on each host, and move the Nutanix CVM startup to Automatic startup. But, according to This VMware KB the automatic startup feature is disabled when moving an ESXi host in an HA enabled cluster. Just like the error message says.

Is this expected behavior? Can this error message be disabled because it is just not valid in combination with VMware HA enabled clusters? Or is this still a configuration error?

This topic has been closed for comments

3 replies

Userlevel 4
Badge +20
The Nutanix setup book also ask you to disable HA for all CVMs, so the configuration goes like this:
- Create the HA cluster
- Add your Nutanix nodes
- Disable HA for the CVMs in "vSphere HA -> Virtual Machine options" by setting "VM restart priority" to "Disabled" for the CVMs.
- Re-enable VM automatic startup at the ESX level for each node.

I currently have this setup and everything is working fine, with the CVM properly booting after a node is powered on.

Sylvain.
Badge +5
When Vcenter does a reconfigure of Vmware HA the setting get's disabled everytime
so the warning will return.
Also, in Vsphere 5.5 webclient I can't even active automatic startup anymore when HA is enabled.
In the 5.5 manual (Manual) you will find this line
"The Virtual Machine Startup and Shutdown (automatic startup) feature is disabled for all virtual machines residing on hosts that are in a vSphere HA cluster. Automatic startup is not supported with vSphere HA."

I opened a case for this at Nutanix and they gave these "workarounds"

We can disable HA, enable the auto-start and then enable HA back again, or we can (besides that) place an script in the ESXi hosts to make sure that our VM is auto-powered up upon reboot.
If you have a look at the instructions provided at the following link: http://download.nutanix.com/guides/c_3_5/xhtml/index.html#oxy_ex-1/topics/vsphere/host_upgrade_complete_vsphere_t.html
In step 3 (Configure the vSphere host) you can find the instructions on how to apply the script to make sure the CVM is powered on automatically on every reboot. In regards to those steps, please be aware of the fact you don't need to reboot the ESXi host to apply the configuration.
As long as you change the script and save the configuration, it should work upon reboot.
You can always do a reboot just to make sure though.



Hope this helps
Badge +5
We discovered this same error and were told that when we go to VMware 5.5u1 the issue is resolved (we like most Nutanix customers are not on 5.5u1 yet because of the NFS bug).

In the mean time, whilst waiting for the upgrade we disabled the alert from the NoS web admin. Its alert A1057.

Please let me know if you ever fixed this issue.