kps79
Contributor
Contributor

HA Failover without a reboot?

Jump to solution

Hello,

I did a few forum searches to discover the answer, but all threads were 4+ years old.  We currently have a EVC Cluster with 3 ESXi 5.0 hosts, running vCenter 5.0 also.  HA works now in the sense that if an ESXi server goes down (power plug pulled) the VM's will migrate and reboot.  Is there a way to set up HA that will migrate the VM without a reboot?  I assume this would take some kind of resource sharing on our storage device, but haven't seen any docs on the subject at VMWare's site.

Thanks in advance for any assistance!

0 Kudos
1 Solution

Accepted Solutions
PduPreez
VMware Employee
VMware Employee

Hi and welcome to the forums

The VMware solution to achieve this is "Fault Tolerance" call FT

FT basically makes a mirror copy of the VM on another host and keep the Memory and CPU in lock step sharing the same disk.

As soon as the primary VM host fails, the FT VM will take over and a new "copy" will be created on another host in the cluster.

See: http://www.vmware.com/products/fault-tolerance/overview.html

There are a few limitations though.

Only supported for 1 vCPU VMs

Seeing your cluster is EVC enabled it might not work in your environment.

EVC downgrades CPU functionality to accommodate all hosts in a cluster.

Please award points if you find this helpful/correct by clicking the tabs below Smiley Happy

View solution in original post

0 Kudos
4 Replies
vmroyale
Immortal
Immortal

Hello and welcome to the forums.

HA will always restart the VMs, if the host fails. That is how it works.

The only way to avoid this would be to use vMotion to migrate VMs before the failure, but that isn't really possible either for unprecited failures. Smiley Happy

Good Luck!

Brian Atkinson | vExpert | VMTN Moderator | Author of "VCP5-DCV VMware Certified Professional-Data Center Virtualization on vSphere 5.5 Study Guide: VCP-550" | @vmroyale | http://vmroyale.com
0 Kudos
PduPreez
VMware Employee
VMware Employee

Hi and welcome to the forums

The VMware solution to achieve this is "Fault Tolerance" call FT

FT basically makes a mirror copy of the VM on another host and keep the Memory and CPU in lock step sharing the same disk.

As soon as the primary VM host fails, the FT VM will take over and a new "copy" will be created on another host in the cluster.

See: http://www.vmware.com/products/fault-tolerance/overview.html

There are a few limitations though.

Only supported for 1 vCPU VMs

Seeing your cluster is EVC enabled it might not work in your environment.

EVC downgrades CPU functionality to accommodate all hosts in a cluster.

Please award points if you find this helpful/correct by clicking the tabs below Smiley Happy

View solution in original post

0 Kudos
TomHowarth
Leadership
Leadership

Hi and welcome to the forums,

firstly in answer to your question,  HA is acting as designed,  when a host fails all the guests on that host are effectively failed too,  what HA does is take thoses failed guests and restart them on the remaining hosts in the cluster.

if you need a seamless failover then you need to investigate FT (Fault Tolerance)  however this is a very resource intensive process, and should be used sparingly,  with FT a guest is kept in lockstep and if the host that runs the primary Guest fails the secondary running on another host will seamlessly take over process,  this guest is an exact replica of the original failed guest, including MAC Adr, IP Adrs, and GUIDs etc.

All that said,  there are some significant pre-reqs for FT the most restictive one being that a FT Guest is limited to a single vCPU.  This requirement does somewhat limit its usefullness.

Tom Howarth VCP / VCAP / vExpert
VMware Communities User Moderator
Blog: http://www.planetvm.net
Contributing author on VMware vSphere and Virtual Infrastructure Security: Securing ESX and the Virtual Environment
Contributing author on VCP VMware Certified Professional on VSphere 4 Study Guide: Exam VCP-410
EdWilts
Expert
Expert

kps79 wrote:

I did a few forum searches to discover the answer, but all threads were 4+ years old.  We currently have a EVC Cluster with 3 ESXi 5.0 hosts, running vCenter 5.0 also.  HA works now in the sense that if an ESXi server goes down (power plug pulled) the VM's will migrate and reboot.  Is there a way to set up HA that will migrate the VM without a reboot?  I assume this would take some kind of resource sharing on our storage device, but haven't seen any docs on the subject at VMWare's site.

I think we're having difficulty understanding your real question...

What I think you might be asking is actually unrelated to HA.

To simply migrate a single VM to another, right-click the VM and select migrate and then "Change Host".

To migrate all of the VMs from one host to another, right-click the host and select "Enter Maintenance Mode".  No VM will run on that host until you exit maintenance mode.

.../Ed (VCP4, VCP5)
0 Kudos