Enter Maintenance mode: Operation Time out

I’ve got this question a couple of time through email and in my training. Let me explain the error (in my case because there can be multiple causes).

A while ago I giving a VMware course where I had 2 ESX host available. Both host where member of a HA en DRS cluster. When I tried to put one host in maintenance mode I took forever and finally I ended up with the error:

Operation Time out

Looking at the Tasks and Events tab didn’t make me any wiser. I looked like a VM was the problem but this VM migrated successfully. The real cause of the problem was HA. In this configuration HA prevented to put one host in maintenance mode because then HA didn’t have enough hosts available for failover. Temporally disabling HA solved the problem.

Again this is one solution. There can me multiple causes witch prevent putting a host in maintenance mode.

About Michael
Michael Wilmsen is a VMware training/consultant (no specific order) with more than 15 years in the IT industry. Main focus is VMware vSphere, Horizon View and Site Recovery Manager with a deepdive to performance. Michael is VCDX 210, VCAP 4/5 certified, has been rewarded with the vExpert title from 2011, is a PernixPro, Nutanix Tech Champion and a Nutanix Platform Professional.

4 Comments to “Enter Maintenance mode: Operation Time out”

  1. By Collin C. MacMillan, January 7, 2011 @ 22:51

    The Adminission Control policy works to prevent you from administratively removing resources that are required to meet the HA policy’s minimum requirements. Changing your HA Admission Control policy to Disable would achieve the same result without totally disabling HA.

    When used in a DRS context, VM’s that violate DRS separation policy will not (automatically) relocate, nor will the VMware-aware applications like vCenter, VUM and VMDR. Such “stuck” VMs will require manual relocation and/or vMotion to the ESX host(s) that is not entering maintenance mode.

    In a two host cluster this issue is irrelevant, but in a 3+ host cluster it’s a better practice. Although the risks of ESX host failure are low, keeping HA in effect will provide better continuity and there’s no risk of “forgetting to turn HA back on.”

  2. By Collin C. MacMillan, January 7, 2011 @ 22:51

    The Adminission Control policy works to prevent you from administratively removing resources that are required to meet the HA policy’s minimum requirements. Changing your HA Admission Control policy to Disable would achieve the same result without totally disabling HA.

    When used in a DRS context, VM’s that violate DRS separation policy will not (automatically) relocate, nor will the VMware-aware applications like vCenter, VUM and VMDR. Such “stuck” VMs will require manual relocation and/or vMotion to the ESX host(s) that is not entering maintenance mode.

    In a two host cluster this issue is irrelevant, but in a 3+ host cluster it’s a better practice. Although the risks of ESX host failure are low, keeping HA in effect will provide better continuity and there’s no risk of “forgetting to turn HA back on.”

  3. By Mike, January 11, 2011 @ 14:02

    Collin, you’re right. This was just a quick solution for the problem in the training. In real live it’s not a best practice to make use of a 2 node cluster. I call this a fail-over situation because you can only use 50% of your total resources in the cluster. 3 or more nodes is always better.

  4. By Mike, January 11, 2011 @ 14:02

    Collin, you’re right. This was just a quick solution for the problem in the training. In real live it’s not a best practice to make use of a 2 node cluster. I call this a fail-over situation because you can only use 50% of your total resources in the cluster. 3 or more nodes is always better.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

You must be logged in to post a comment.