asiajason
Contributor
Contributor

VMHosts don't restart in HA Cluster upon failure of host

Hi,

We have two hosts, ESXi v4.1 (vCenter 5), HA clustered.  One host shuts down periodically, and the VM's are migrated to the to the other host, but they don't automatically start up.  I thought in properly configured HA, vm's are powered back on another host in the cluster automatically.

My cluster settings:

vSphere HA is checked

DRS is NOT checked

Host monitoring is checked

Admission control is disabled

vm options:

vm restart priority is High

Host isolation response is "shutdown'

-- i suspect this option, but i've found this is correct option.

-- vSphere HA isolation response... which to use when? - VMware vSphere Blog

Any ideas on how to get hosts to restart automatically?

Thanks

Jason

Tags (1)
0 Kudos
10 Replies
vijayrana968
Virtuoso
Virtuoso

Probably this was detected as host isolated, for which you have set policy to shutdown ! Could you dig up logs more at location /var/log/vmware/aam on host.

0 Kudos
asiajason
Contributor
Contributor

Hi,

We don't have any recent logs in that directory - last logs are 2013...

Regarding shutdown option, according to this article:

vSphere HA isolation response... which to use when? - VMware vSphere Blog

"Choose shutdown to allow HA to restart VMs on hosts that are not isolated and hence are likely to have access to storage"

If I choose to not shutdown, the vm's will continue to remain on in the isolated host? - and in our case, that host is having a memory problem, so it means the hosts may not be available. 

Ideas?

0 Kudos
asiajason
Contributor
Contributor

Any other suggestions?  I'm not expert, and don't know why there are no logs in the directory you specified.

I'm happy to follow any leads you have for me.

Thanks!

0 Kudos
vijayrana968
Virtuoso
Virtuoso

Can you post cluster's each HA setting with snippet image.

Also make sure if host wasn't isolated which you have configured response to shutdown VMs. Understanding High Availability Host Isolation Response with Network Attached Storage (1018325) | VM...

0 Kudos
asiajason
Contributor
Contributor

Hi, here are the cluster settings:

pastedImage_0.png

pastedImage_1.png

pastedImage_5.png

pastedImage_6.png

pastedImage_7.png

pastedImage_8.png

pastedImage_9.png

0 Kudos
vijayrana968
Virtuoso
Virtuoso

What are settings for virtual machine options.

0 Kudos
dbalcaraz
Expert
Expert

Hi,

Hi,

We have two hosts, ESXi v4.1 (vCenter 5), HA clustered.  One host shuts down periodically, and the VM's are migrated to the to the other host, but they don't automatically start up.  I thought in properly configured HA, vm's are powered back on another host in the cluster automatically.

If you could not find logs in the esxi hosts, could you check logs from vCenter? The most helpful would be the vxpd.log file.

Also, can you share your vSwitch configuration?

Are trying to get any clue of why this host is restarting or only you want to know how VMs aren't restarting in other hosts?

Best,

Dan

-------------------------------------------------------- "I greet each challenge with expectation"
0 Kudos
Punitsolanki
Enthusiast
Enthusiast

The minimum requirement for HA is 3 hosts, hence it will not work as expected.

Punit Solanki psolanki@vmware.com
0 Kudos
RajeevVCP4
Expert
Expert

Please provide logs

  • /var/log/vmware/aam
  • what was the date and time
  • Any virtual machine vmware.log which you rebooted manually.
Rajeev Chauhan
VCIX-DCV6.5/VSAN/VXRAIL
Please mark help full or correct if my answer is use full for you
0 Kudos
vijayrana968
Virtuoso
Virtuoso

Since you dont have relevant AAM logs, next step would be to check vxpd.log logs.

VMware has HA simulator the analyse the predictive HA action in case of host failure and you can test it at http://hasimulator.vmware.com/html/index.html#/summary

I am not sure this can check your cluster settings as it DRMDUMP files which is generated with DRS which you haven't enabled.

DRMDUMP : This log file been generated once your HA configures and updates once any changes made. Which includes Actions proposed and taken by VMware Distributed Resource Scheduler (DRS), grouped by the DRS-enabled cluster managed by vCenter Server. These logs are compressed.

Help resource : VM Resource and Availability Service

0 Kudos