Possible to specify IP used for mounting NFS datastore

[PTide]

Hi,

Your post has been merged into the existing discussion.

As to your problem - was mount operation finally successful or it just stopped there? The thing is that since 9.5 U2 VBR server passes all available IPs and hostnames to ESXi host and tries to mount NFS using one of those addresses/names. If it fails it should try the next on and so on. If that's not the case then please contact our support team for further investigation.

Thanks

[Chri]

Hi

No, it was not successful. It just stops

[PTide]

Therefore, it is an unexpected behaviour and I'm looking forward to you posting your support case ID

Thank you

[Chri]

Hi

I think i found the mistake:

As for KB1055 another process was locking port 111
I think it was the Windows NFS service, which we also use on one of our repositories.

So i change the mount server of the repository to the server where no Windows NFS service is running.

Now the FLR mount seems to be working

Thanks

[balma01]

Vmware vsphere installed on a Nutanix HW.
I have installed a proxy for each Nutanix node, so the proxy can acces the NFS server local IP (192.168.5.2) directly and not across the network. Every proxy is dual homed, the first NIC on the internal Nutanix vSwitch has a 192.168.5.x IP address and it's used to connect via direct NFS the NFS server, the second NIC ip is a 192.168.100.x and it's used for the communication to the VBR server.
Since the repository is a data domain, a proxy (proxy9) is performing the gateway server and the NFS mount server roles.
When I try to do an instant restore in host1, the host tries to mount the NFS datastore using proxy9, hosted in host9, as mount server but the NFS mount fails because the host is pointing to the 192.168.5.9 IP that is the proxy9 internal IP.
How to force the hosts to use the proxy external IPs to do the NFS mount ?

[foggy]

Hi Mauro, mounting the datastore manually should help - see this thread for details. Thanks!

[rurouni]

Hi,
I'm facing the same issue on vsphere proxy on Nutanix nodes.
Randomly, the internal Nic dedicated to nutanix traffic is used to mount the Vpower NFS Datastore
However it is not reachable which is by design. Is it possible to specify on which Nic the NFS datastore will be mounted?

Thanks

[foggy]

Have you tried to manually mount the datastore using the address in the required network (see above)?

[tyler.jurgens]

14 years later - is this still the best way to force the NFS mount over a specific network?

Just to add a bit of colour to this - I added a single ESXi host directly to my VBR server via the NFS network I wanted to use, for performing an instant VM recovery. Veeam then mounted the NFS share to the ESXi host via the mgmt network of the VBR, not via the NFS IP.
I would have expected Veeam to try mounting via that network (hence why I added it directly via that network).

I want to do an instant VM recovery and it chose the least efficient network to do so by default. Would be a HUGE benefit if I could specify the network to use within Veeam itself.

[tyler.jurgens]

Also, I was unable to mount that share directly to the ESXi host using the above mentioned process. When I create an instant VM recovery, it seems to create the share at that time. When I tear down the instant VM recovery, the share no longer exists, so I cannot mount to that NFS share manually via the correct network (using IPs for everything, not DNS names).

To work around this issue I disabled the management NIC on the VBR, leaving the NFS NIC the only NIC that Veeam could present a mount on. This works to mount the NFS share via the proper IP to the ESXi host.

Not as instant as I would like. Please add the feature to specify the network to use to mount the NFS shares.