Instant Recovery fails to mount vPower NFS datastore

[AALSG]

I was trying to test Instant Recovery, however, i could not get it working.

I am not sure what i am missing.

Here are my steps:

1. Full Backup VM using veeam
2. Delete VM
3. Activate Instant Recovery, error occurred.

There was an error saying that it was unable to mount the vPower NFS datastore.

I am on vSphere 4.1.

Any one have the same error as me?

[Vitaliy S.]

Hello,

It's really hard to say what might be the reason without seeing full log files. Please contact our technical team and provide them with full log files for investigation.

Thanks!

[Alexey D.]

You may also take a look at this topic: Unable to mount vPower NFS volume

[AALSG]

Hi all,

Thanks for the response, i did some investigation and found out that the person who configured the VMware for us put the vmkernel port group on a separate subnet. The issues is resolved when i created a VMKernel port group in the same subnet on the machine where Veeam is installed.

How can i connect the NFS Volume without creating the new vmkernel port group?

[Gostev]

Hi, unfortunately this is not possible with VMware hypervisors... ESX(i) host can only access NFS storage through vmKernel interface. Thanks!

[AALSG]

Sorry, i think i did not put my question right.

What i meant was how can i mount the NFS volume on the existing vmkernel which is on the different subnet without creating a new vmkernel on the same subnet as the Veeam VM.

[Gostev]

This should work out of box if you have routing setup correctly between your subnets? Do you?

[AALSG]

Hi Gostev,

You are right, i checked with the vendors and they mentioned that there wasn't any routing configured for those subnets to route.

[Gostev]

Sounds good, thanks for taking time to update the topic with confirmation.

[AALSG]

The routing have been configured and everything works as it should! Thanks for the help!

[GabriMWare]

Here the solution!
In my case, I've been failing in the vSwicth configuration (on VM Ware). First of all, you have to add a VMKernel port on the same vSwitch where you have the VM portgroup (in order to mounting NFS system). Then, configure VMKernel port with an IP address using the same subnetwork where is placed the server where is Veeam installed.
In my case, the mistake was that I've added the VMKernel port to the vSwitch, but it was configured using another subnet (range).

Then, the NFS system has been mounted and veeam has showed me the browser with Linux files.
Great!

[Gostev]

Thanks. May be we should add automated vmKernel configuration, to prevent these sort of mistakes...

[zak2011]

I was trying to test out Instant Recovery and it did not succeed unfortunately.
I got the error 'Unable to mount the NFS datastore error'.
The VM kernel interface is on the same subnet as the Veeam server and i am able to ping the Veeam server from the console of the ESX host.
In my case we have added the VMkernel interface on a new vswitch and assigned it an IP with the same subnet as the veeam server. Is this correct?

[Vitaliy S.]

Could you please tell me if you're able to mount vPower NFS datastore manually?

[tsightler]

Well, I'm not sure about the statement that the VMkernel is on a new vSwitch. Is that vSwitch connected to the same layer2 network as the Veeam server? If not, then this is not correct. Have you tested simply pinging the VMkernel IP address from the Veeam server?

[zak2011]

I tried doing a vmkping from the host and it is not able to ping the Veeam server. So, i guess i wont be able to mount the datastore manually either.
What would i be doing wrong ?

Thanks!

[tsightler]

If I understood your original post you indicated that you created a new vSwitch but used an IP address from the same subnet as the Veeam server. If this vSwitch is not connected to the same layer-2 network as the vSwitch your Veeam server is on, then this will not work. I'd try simply creating a VMkernel interface on the same vSwitch has your Veeam server.

[zak2011]

Thanks Tom. That seems to be most likely be the reason. The only reason why the VM admin was hesitant to create the vmkernel inteface on the same vswitch as the Veeam server was not to mix both the traffic.
Right now we cannot use vmotion anyways due to our license limitation. So it shouldn't be a cause of concern creating the vmkernel interface on the same vswitch as the veeam server, isnt it?

[tsightler]

For most small environments this is usually what I see anyway and it shouldn't cause a problem. If you want it to stay on a separate vSwitch then you would need to add a second vNIC to the Veeam server and make sure that vNIC is also attached to the "new" vSwitch. Then assign IP addresses from a different subnet to the VMkernel and secondary NIC in the Veeam server.

[zak2011]

Thats cool!. I tried creating the VMkernel on the same VSwitch as the Veeam server and it worked. Instant Recovery works like a breeze.
I guess i dont need to create a seperate it at the moment as i dont have any vmotion traffic.
Thanks a lot Tom.