VMware FT has an issue in starting secondary VM

[Availability]

I have been able to enable FT on a VM which is shared on NFS meeting all the prerequisites of 1Gbps NIC support. The ESX hosts on which I am enabling FT are also compatible as confirmed by the Site Survey analysis. The Fault tolerance logging has also been enabled on a NIC and vMotion on the other NIC.

I have also been able to Turn on Fault Tolerance on a VM, but as soon as the secondary VM creation task is completed, FT is disabled on primary VM stating that the secondary VM could not be powered on. The error details state that "Virtual Machines in the same Fault Tolerance pair cannot be on the same host". And for the other server, it states "Operation Timed Out"

We have ensured that primary and secondary VM are not on the same host. Please find the screenshot attached.

Please respond.

[krowczynski]

Hmmh ,

looks strange, what abour you HA Config, is there everything configured properly?

MCP, VCP

[Availability]

Yes HA has been correctly configured.

[krowczynski]

OK, have you located any error in the logfiles, maybe?

MCP, VCP

[Availability]

Still checking the log files but there is one more thing that I have noticed that in spite of the fact that the hosts in the FT cluster have the same build and vMotion can successfully happen between these, the error log gives me the following error:

"No compatible host on secondary VM FT_Test2. "

Please find snapshot for more details.

[krowczynski · Accepted solution]

Well,

what ybout your both hosts in the Cluster, are all of them fulfill all requirement for FT?

Also read this one:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100895...

MCP, VCP

[Availability]

Yes, the hosts fulfill all the requirements for FT as ensured from the site survey analysis.

Also it seems that vmotion is timing out. Although if we vmotion separtely, it works fine. Please find the log file attached.

(1261117624584894) VMotion failed: vim.fault.Timedout

Failed with: vim.fault.PowerOnFtSecondaryFailed

migration errors:

fullFormattedMessage = "Message on FT_Test2 on 192.168.111.84 in DC: Error encountered while trying to restore the state of group StateLogger during migration.

Failed to receive migration.

",

changeTag =

[krowczynski]

What abou you network config, if vmotion time out, maybe there is something wwrong.

Have you got the same portgroup for vmotion and ft or have you separated this.

MCP, VCP

[Availability]

I have tried vMotion on the same cluster, but that works correctly. And I have ensured that vMotion and FT are on separate Port groups.

[krowczynski]

Well, the last thig I would try is to disable the FT on this portgroup, disable the whole cluster and enable it back again, and create a new portgroup for FT.

I know it is only for trying, but mybe it helps.

MCP, VCP

[Availability]

One more thing, there is a pre-requisite for Jumbo frames for FT. I tried enabling Jumbo frames on the ESX host through this:

esxcfg-vswitch -m 9000 vSwitch1

and I got the following error message:

Unable to set MTU to 9000 the following uplinks refused the MTU setting:vmnic1

FYI: name of NIC is NetXtreme Broadcom 57...

Do you think this could be a possible reason?

Also how do I ensure if my ESX host has hardware virtualization enabled in BIOS, as per the requirements for FT ?

[krowczynski]

>>Do you think this could be a possible reason?

Maybe it could be possible.

>>Also how do I ensure if my ESX host has hardware virtualization enabled in BIOS, as per the requirements for FT ?

Go under BIOS and look in the CPU Menu, if you see something like virtualization.

MCP, VCP

[Availability]

Thanks, reboot and disabling and re-enabling HA helped. In fact no change in setting was required.

[krowczynski]

Thats sounds great!!

• Please award points if answer was cooecr or helpful ***

MCP, VCP