I cancelled a vmotion task and it is now stuck in "Cancel Requested" status

[syt75013]

Hi everyone,

I've cancelled a Storage vMotion + vMotion task on a VM yesterday and now the task is stuck in a "Cancel Requested" status.

I am now unable to manage the VM as the actions are greyed out(power off, migrate, snapshot, ect) but it's still responding.

Since the last 24 hours I am waiting for the task to timeout but it's still stuck in this state "Cancel Requested" ...

Have you guys encountered a similar issue?

Thanks !

[Jonathan77]

You can try to stop the task using powerCLI..

This worked for me in the past to stop hung task...

If the task is still running... something like :

To see running tasks and find its task ID:

Get-Task -Status "Running" | select Name, Id, State

To stop the task:

(replace "task-ID" with the one located in previous command)

 Get-Task -Id "Task-ID" | Stop-Task -Confirm:$false

 Jonathan P.

[AveryFreeman]

Did you ever resolve this issue?

I cancelled a simple migration from one datastore to another and it hung, leaving the source datastore in limbo. 

Now I get the same error message both in vCenter and through powercli: "The operation is not allowed in the current state." which is rather cryptic, and I don't seem to be able to get any more information than that.

I tried to list the source datastore in SSH or through the web UI and both are inaccessible.  I'm afraid I just destroyed an entire VM by simply hitting the cancel button.

[maksym007]

Another option to kill the task directly on ESXi host via SSH. 

[AveryFreeman]

I started the task in vCenter, so I actually powered off the vCenter VM and rebooted the host (it's a new one I'm working on setting up still).

Lucky for me, the VM was in tact, it just had compute and storage resources on two separate drives.  I migrated the storage back to its original location, tested it and it was fine.

I don't recommend this as a solution, it's likely nobody else will be as lucky as I was.

[esximan123]

Me too. I powered off the vCenter VM and rebooted the host, prayed. Then when it all came back online I deleted some files that were on the host I was trying to migrate to. Then turned on the VM on the original host and by the grace of God it turned on. Will not be doing that again..

I also found that for some reason migration when the VM is turned on is way faster then if it was off. No idea why I have 10Gig but when the VM was off it transferred at 900Mb/s and when it was on it transferred at 6,500 Mb/s

[a_p_]

You may or may not have noticed that you were replying to a ~2 years old discussion.

Anyway, please read https://core.vmware.com/resource/vsphere-vmotion-unified-data-transport, which explains the performance difference, and what has been changed with vSphere 8.0 to speed up storage migrations.

André

[syt75013]

Hi,

If I remember correctly, I resolved the issue by restarting the vpxa and hostd service on the source ESXi

/etc/init.d/hostd stop

/etc/init.d/hostd stop

/etc/init.d/vpxa stop

/etc/init.d/vpxa start