Unable to send backups to VBR server

[dgomes]

Hello,
During the beta for VAL there was a known issue for sending backups to VBR. We are now re-testing with VBR 9.5U1 and the latest VAL on an ubuntu system hosted at a cloud provider.

We seem to get further than during the beta but still having an issue. The VAL or VBR server (not clear which) is pointing to the "wrong" place.

Code: Select all

16:18:37        Starting full backup to [svr-bkp-01]repo-01
16:22:57        [error] Failed to perform backup
16:22:57        [error] Failed to connect to the port [192.168.60.31:2500]

The complete data path is as follows:

[Cloud ubuntu] --> [firewall NAT @ customer lab] --> [VBR server] --> [Create Backup Data]

The IP address 192.168.60.31 is a valid IP for the VBR server, but this is a secondary NIC on a closed network for storage (iSCSI). By observing firewall logs I see the initial connection works fine and a small backup file of a few KB (.VBM) is even created in the assigned backup repo, however soon after the job fails with the above error message.

The iSCSI storage is mounted at the windows level and is visible as a local drive letter to the VBR server, and is configured as local storage in VBR (windows server), so there is no need for veeam itself to be trying to access this network.

I've tried creating an entry in the VBR server's hosts file for the correct IP for the machine name (svr-bkp-01) but this did not help.

Are there any specific configurations required for this to work?
If required I can test with a NAS-based repo but the customer would prefer the backups all go to his iSCSI device.
Thanks,

[PTide]

Hi,

Please check "Used ports" section in system requirements and make sure that the firewall has all required ports opened.

Thank you.

[dgomes]

Hi,

Please check "Used ports" section in system requirements and make sure that the firewall has all required ports opened.

Thank you.

Hi, All those ports are NATed from WAN to the correct internal IP of the backup server (192.168.5.245). However all the error messages/logs talk about failing to connect to 192.168.60.31:2500. It is normal that this fails because the 192.168.60.x network is not routed, it is a storage network, the VBR server's second NIC is configured on this network for windows to map iSCSI devices. From my point of view it looks like VBR is giving the wrong NIC to VAL.

My next test will be to set up a temporary VBR server on a standalone machine with a local disk and only 1 NIC and I suspect this will work.

Thanks,

[PTide]

From my point of view it looks like VBR is giving the wrong NIC to VAL.

Yes, you are right - a similar issue can be seen when using Windows Endpoint with VBR that has multiple IPs.

My next test will be to set up a temporary VBR server on a standalone machine with a local disk and only 1 NIC and I suspect this will work.

Creating a NAT rule on the VAL host that forwards packets with destination 192.168.60.31 to 192.168.5.245 should also work.

Thanks

[abaldinazzo]

Hello everybody,
I have a very similar problem: a Veaam Backup & Replication server (SRV-BCK) on a network (IP: 192.168.20.244) that comunicate with a Veeam Backup Repository (SRV-VBR) on another network (IP: 10.0.0.245). They communicate over an OpenVPN, so the server SRV-VBR has IP 172.16.0.1 and the server VBR-BCK has IP 172.16.0.10.
When I start a Backup Copy Job from SRV-BCK to SRV-VBR I get the error "Failed to connect to the port [10.0.0.245:2500]": obviously, because the network 10.0.0.0 is not visible from SRV-BCK.
Can you please explain how to create the NAT rule that should solve the problem?
Thank you!