Possible to specify IP used for mounting NFS datastore

[lewisdrummond]

My Veeam Backup machine has muliple NICs and I can't seem to specify which IP address should be used when trying to create a Virtual Lab. The Lab creation fails when trying to mount a datastore because the ESX box can only access one of the Veeam Backup IP addresses.

Any advice?

Regards,

Lewis

[Gostev]

I am looking into this for you.

[Gostev]

Try to add NFS datastore to selected ESX host manually before creating virtual lab.

1. vSphere Client, ESX, Add Storage, NFS.
2. Server: enter IP address or DNS name of Veeam Backup server. ESX has to have vmKernel on the corresponding interface.
3. Folder: enter /VeeamBackup_SERVERNAME (where SERVERNAME is NetBIOS name of Veeam Backup server).
4. Do NOT select Mount NFS read-only checkbox (leave this one unchecked). vPower engine will handle disk writes correctly by itself.
5. Datastore name: enter VeeamBackup_SERVERNAME (no forward slash here).

Then try again.

Whether or not this will help, please shoot us VeeamBackupService.log from Help > Support Information to support@veeam.com referencing this thread. Our NFS logic is designed to handle presence of multiple network adaptors on Veeam Backup server, so we do want to know what went wrong anyway.

[Gostev]

I think what happens is resolving Veeam Backup server name on ESX host side returns "wrong" IP address (address from the network this ESX has no access to).

[lewisdrummond]

The ESX box was resolving the correct IP address. I was able to manually create the NFS mount and it seems to work. Thank you.

[Gostev]

Sounds great! If it is not too much to ask, we would really appreciate if you could still sends us VeeamBackupService.log log file, so that we have better understanding of what has happened. Thank you!

[lewisdrummond]

Where do you want me to send it? Full path to log?

[Vitaliy S.]

Hello Lewis, you may locate the log file required if you click Help->Support Information in the backup console. Please send it to support@veeam.com with the reference to this topic as Anton has asked in his second post above. Thanks!

[lewisdrummond]

Done!

[afokkema]

Got the same problem: Unable to mount vPower NFS volume. I will send the logfile to support@veeam.com.
Got a Windows 2003 VM with two network adapters. An internal LAN and a vSphere LAN. I cannot mount the share directly from a vSpher host.

[Gostev]

Thanks, this will help us in researching this issue.

[afokkema]

Fixed the issue. The vSphere host couldn't reach the Veeam server via VMKPING. So the vmkernel network was not in the same subnet like the Veeam server. So I added a new vmkernel network in the correct subnet. See my blogpost: http://ict-freak.nl/2010/10/25/veeam-in ... ilesystem/ for more details.

[Gostev]

Thanks for the update, Arne. Hope you enjoyed working with our support!

[snrpepie]

I am having possibly the same issue, however a manual mount did not seem to affect the outcome.


kernel: 257:23:43:46.838 cpu8:1064)WARNING: ScsiDevice: 3264: Failed for vml.02000e000060050763054e000000000000010a323130373930: Transient storage condition, suggest retry
Nov 30 10:57:08 esxhost vmkernel: 257:23:43:46.838 cpu8:1064)WARNING: ScsiDevice: 3633: READ CAPACITY on device "vml.02000e00006005076305fe000000000000010a323130373930" from Plugin "legacyMP" failed. Transient storage condition, suggest retry
Nov 30 10:57:08 esxhost vmkernel: 257:23:44:57.293 cpu9:1065)NFS: 107: Command: (mount) Server: (12.34.56.78) IP: (12.34.56.78) Path: (/VeeamBackup_BACKUP) Label: (VeeamBackup_BACKUP) Options: (None)
Nov 30 10:57:21 esxhost vmkernel: 257:23:45:28.291 cpu9:1065)WARNING: NFS: 898: RPC error 13 (RPC was aborted due to timeout) trying to get port for Mount Program (100005) Version (3) Protocol (TCP) on Server (12.34.56.78)

Any help would be appreciated. I am trying to do a file level restore of a Solaris VM, it just bombs..

thanks!!

[Gostev]

Does your ESX host have vmKernel interface on the network with Veeam Backups server? Can you vmkping the Veeam Backup server IP? If your Veeam Backup server has more than one NIC, does DNS lookup from ESX server return expected IP address? Do you have any firewall on Veeam Backup server, or between ESX and Veeam Backup server? Is vPower NFS service running on Veeam Backup server?

If you are unable to pinpoint the issue with the above checks, please open a support case and provide full logs for investigation.

[snrpepie]

my esx host does not seem to be able to resolve the hostname of the veeam server. How can I remedy this?

thank you

[Gostev]

Check ESX host's DNS server settings, or DNS itself?

[BarryCoombs]

Hi

I am installing Veeam B&R for a customer at present, the customers production network is only 100MB and the iSCSI network is obviously 1GB, I notice that by default the Veeam datastore mounts to the ESX host via name and presumably the production LAN.

I need to get the NFS datastore for instant recovery etc mounted over the iSCSI network that both the backup and ESX hosts have access too, worst case I will need to get it mounted over a legacy backup lan that is also 1GB.

What is the best practise for this?

Cheers

Barry

[BarryCoombs]

So to confirm I need to add the Veeam server nfs datastore manually via ip of the iscsi interface on the Veeam host? My Veeam host has 4 iscsi interfaces do I just need to pick one rather than getting the benefit of all the nics?

[Gostev]

Correct. vPower NFS does not have MPIO software just yet

[BarryCoombs]

Brilliant, Many Thanks Gostev

[NetWise]

In our case, Veeam is using our 10GbE NIC's to hit the SAN for backups, which is working great. I fully understand the inability to write to VMFS directly via SAN. A question though - is there a way to tell the vPower when it presents the NFS to the servers, to use the 10GbE interface(s)? This would make my NFS mounts much quicker and the associated svMotion back to the SAN as well.

[StefanSpecht]

Hi all,

I trying to find the best way to force vPower to connect its NFS datastore using the vmkernel-port for storagenet.

ESXi host has
-> vmk0 for management network
-> vmk1 for nfs network

vPower NFS server has
-> Teamed NIC connected to management network
-> Teamed NIC connected to storage network

If I perform a instant recovery, vPower always uses the management IP of the host to provide its datastore. How to tell him to use nfs IP?

Regards,
Stefan

[veremin]

Your post has been merged into the existing discussion.

As to your problem, it should be solved by manual adding vPowerNFS datastore to your ESX(i) host via storage network. Please take a look at this particular post for the full description of the process.

Hope this helps.
Thanks.

[vsenguel]

Hi, i have here a VBR Hardware Server with 3 nics (1Gig, 2 x 10gig).

I would to configure vPower to deliver the instant recovery over the same Network like the ESXi Hosts for the best Speed because the ESXi Hosts (vCenter managed) Management Network are all connected via 10gig.

How can i set the primary ip for vPower NFS service?

thank you for any help
Volkan

[erth111]

Hi guys,
I'm deploying Veeam B&R in vmware environment with multiple hosts and I've stumbled upon the problem of the interface that is used for mounting vPowerNFS on the ESXi host. First of all, why there is no information about the connection being established from ESXi host to the Veeam server in "Used Ports" section of the User Guide? Second, in my opinion, the method of forcing the ESXi to mount NFS datastore over a specified ip adress by manually mounting the datastore is not flexible in big environments.

Is there any other way to specify a certain IP address to be used to mount the NFS datastore?
What if i change the DNS record of veeam backup server so it points to the desired ip address? Will this work?

[Vitaliy S.]

Hello Tomasz,

I'm deploying Veeam B&R in vmware environment with multiple hosts and I've stumbled upon the problem of the interface that is used for mounting vPowerNFS on the ESXi host.

Veeam B&R mounts NFS service via VMkernel interface of the ESXi host or are you referring to Veeam Backup server NIC?

Second, in my opinion, the method of forcing the ESXi to mount NFS datastore over a specified ip adress by manually mounting the datastore is not flexible in big environments.

Can you please elaborate on this? Just to make sure I understand all use cases, possibly might have missed some.

Is there any other way to specify a certain IP address to be used to mount the NFS datastore?
What if i change the DNS record of veeam backup server so it points to the desired ip address? Will this work?

Yes, I believe it should do a trick. Please keep us updated on your findings.

Thank you!

[erth111]

Hello Tomasz,

I'm deploying Veeam B&R in vmware environment with multiple hosts and I've stumbled upon the problem of the interface that is used for mounting vPowerNFS on the ESXi host
Veeam B&R mounts NFS service via VMkernel interface of the ESXi host or are you referring to Veeam Backup server NIC?.

I'm referring to Veeam Backup server NIC. There is no way to specify it. My Veeam server, as well as ESXi hosts have NICs in two networks: one for management and one for backup. I would like to able to specify the backup interface to be used for NFS.

Second, in my opinion, the method of forcing the ESXi to mount NFS datastore over a specified ip adress by manually mounting the datastore is not flexible in big environments.
Can you please elaborate on this? Just to make sure I understand all use cases, possibly might have missed some.

Let's say I want to create additional Virtual Lab on different host. Now, the backup administrator has to configure the vmware platform istead of just Veeam console. What about configuration backup? The part that is configured on vmware side is not included in Veeam configuration. Wouldn't it be nice to have an option to specify what interface the NFS service is listening on? Then the ip address of that iterface would be passed to ESXi host to be used as NFS server ip address.

Is there any other way to specify a certain IP address to be used to mount the NFS datastore?
What if i change the DNS record of veeam backup server so it points to the desired ip address? Will this work?
Yes, I believe it should do a trick. Please keep us updated on your findings.

Thank you!

Unfortunalety, I will not have a chance to test it any time soon.

[Vitaliy S.]

I'm referring to Veeam Backup server NIC. There is no way to specify it. My Veeam server, as well as ESXi hosts have NICs in two networks: one for management and one for backup. I would like to able to specify the backup interface to be used for NFS.

Yes, there is no way to specify it in the backup console, since all the routing is controlled by the Window Server where you've installed your backup server.

Let's say I want to create additional Virtual Lab on different host. Now, the backup administrator has to configure the vmware platform istead of just Veeam console. What about configuration backup? The part that is configured on vmware side is not included in Veeam configuration. Wouldn't it be nice to have an option to specify what interface the NFS service is listening on? Then the ip address of that iterface would be passed to ESXi host to be used as NFS server ip address.

Yes, it would be a nice to have feature, I agree, though the DNS trick you've suggested should also work.

[Chri]

Hi

This is our current Veeam environment:

1 Veeam B+R Server with an ip 172.16.x.x
2 Veeam backup repositories with an ip 172.16.x.x
3x VMware servers with an ip 172.16.x.x

The two backup repositiries (Server 2012 R2) have a second network (192.168.x.x), wich is direct attached between these two servers to do some other copy and replication jobs (not Veeam related).

My Problem now is:
i tried to do a Linux guest file restore and the following error message comes up.

Code: Select all

Failed to mount restore point.
Linux FLR appliance deploy failed: Failed to mount NFS
volume (192.168.x.x:/VeeamBackup_172.16.x.x).
192.168.x.x: Fault "PlatformConfigFaultFault", detail "NFS
mount 192.168.x.x:/VeeamBackup_172.16.x.x failed:
Unable to connect to NFS server."
An error occured during host configuration.

So it seems that it is trying to mount it at the direct attached network.
But why? I have never populated this IP to the whole Backup environment.

Thanks