Enter Mainenance Mode - Operation times out when VMs powered on - HA issue?

[RichyM]

Hello,

I want to update our two esx 3.5 server hosts with update manager running on vCenter 2.5. (Mamaged through Infrastructure Client 2.5)

On both hosts vm's are powered on. When I start the remediation on one host an error occurs (Operation timed out). When I configure the failure response to "Power Off and Retry" the host powers off all vm's and the update works.

My question is: Shouldn't the host use HA to power off the vm's, move them to the other host and power them on again? Or is vMotion necessary for this operation?

[athlon_crazy]

+My question is: Shouldn't the host use HA to power off the vm's, move

them to the other host and power them on again? Or is vMotion necessary

for this operation?+

1) Why would the host use HA? Is there any host isolation or even host failure? The answer is no!

2) You should move the vm's to another host via vMotion for plan maintenance or shutdown.

System Engineer

Zen Systems Sdn Bhd

Malaysia

www.no-x.org

[RichyM]

+ 1) Why would the host use HA? Is there any host isolation or even host failure? The answer is no! +

There is no host isolation or host failure but the request for mainenance mode. Since there can't be powered on any vm i thought this is enough to activate HA and move the vm's to the other host automaticallly.

When vMotion is lisenced and activated will the movement of the vm's to another host happen automatically when entering maintenance mode? Will the esx host server update be fully automated without any downtime of the vm's?

[athlon_crazy · Accepted solution]

Forgot to ask you one thing, have you play around with HA setting? (Leave machine Power On, Power off, shutdown). This could make a different result!

Should moving the vm's automatically with vMotion & DRS

System Engineer

Zen Systems Sdn Bhd

Malaysia

www.no-x.org

[admin]

In order VMs to be moved automatically, you should enable vmotion on both hosts and the hosts should be part of a DRS cluster.

[Troy_Clavell]

as others have stated to utilize the automation of VUM, you must have DRS enabled on your cluster. During the remediation, the host will need to be in maintenance mode. VUM issues the command, migrates all the VM"s off the host via DRS, puts the host into maintenance mode, patches the host, reboots the host, and then takes host out of maintenance mode. This process is repeated in the cluster if you have your baseline set on the cluster level.

I have seen manual interaction with VUM sometimes breaks the remediation process. So, if you don't have DRS in your cluster, you may see it breaks the automation of patching.

Hope this helps.