Virtual Lab: routing between two isolated networks

[Ximera]

Hello!

My problem is quite common but I do not have a solution.
I created Virtual Lab with Proxy Appliance. You can see its configuration below:

Code: Select all

Appliance:	
   Name:		Virtual_Lab_tst
   Pool name:	Virtual Lab tst
   Folder name:	Virtual Lab tst

   Production network name:VM Network
   IP:                     10.22.34.218
   Subnet mask:            255.255.255.192
   Default gateway:        10.22.34.193

   DNS:
      Preferred:	10.22.34.198
      Alternate:	10.22.22.3

Network configuration type:	Advanced
Network options:
      Isolated network:  Virtual Lab tst VM Network
      Masquerade IP:     10.255.34.192
      Appliance IP:      10.22.34.193
      DHCP:              enabled
      DNS:               10.22.34.198

      Isolated network:  Virtual Lab tst VM Service
      Masquerade IP:     10.255.229.0
      Appliance IP:      10.22.229.1
      DHCP:              enabled
      DNS:               10.22.34.198

Network mapping:
   VM Network       --> Virtual Lab tst VM Network
   VM Service    --> Virtual Lab tst VM Service

Static mapping:
   10.22.34.194    --> 10.22.34.228   	
   10.22.229.2     --> 10.22.229.77   

As you can see I nave two production networks: 10.22.34.192/26 and 10.22.229.0/25. In production there's bidirectional communication between these subnets.
My Veeam B&R server and Proxy Appliance is located in 10.22.34.192/26 subnet.

When I start to deploy from backup and test VM with address in 10.22.34.192/26, I have no problem with it: I ping masquerading IP from B&R Server, I can get access to VM by static mapping IP. Ping test is also successful.

But when I deploy VM from second subnet 10.22.229.0/25, I could see it neither by masquerading IP nor by static mapping IP. Ping test fails.

B&R Server routing table during SureBackup Job execution:

Code: Select all

Active Routes:
Network Destination        Netmask          Gateway       Interface  Metric
          0.0.0.0          0.0.0.0    10.22.34.193    10.22.34.232    266
    10.22.34.192  255.255.255.192         On-link     10.22.34.232    266
    10.22.34.232  255.255.255.255         On-link     10.22.34.232    266
    10.22.34.255  255.255.255.255         On-link     10.22.34.232    266
    10.255.34.192  255.255.255.192    10.22.34.218    10.22.34.232     11
     10.255.229.0  255.255.255.128    10.22.34.218    10.22.34.232     11

I've read in HelpCenter "If you assign to the proxy appliance an IP address from the network other than that where the Veeam Backup server is located, you will have to manually add a new route to the routing table on the router in the production network. " - I do not fully understand is this a point or not, and if so what route should I add on my core-router?

Any help, manuals, links is appreciated. Thanks in advance.

[Ximera]

I've partially solved my problem: now I can access VM in 10.22.229.0/25 by it's masquerading IP 10.255.229.0/25. But Static Mapping still doesn't work. Is it possible to set up static IPs for VMs in second isolated net?

[foggy]

Hello, Julia, here's a couple of useful links for you:
SureBackup Design Question
Virtual Lab Routing Engine issue

[Ximera]

Thanks for you answer, foggy.

I read these topics, but as I see the point is in distributed architecture: VBR Server and Proxy Appliance are in different subnets.

In my case VBR Server and Virtual Lab are on the same production network, VM on this network is also accessible by mapped IP and by masquerading IP.
Moreover, now VM on second production network is also accessible by masquerading IP from first network - I can ping it or connect using RDP.
But mapped IP is more convenient way to get access to VM and it doesn't work. So I don't know is this behaviour by design or not.

I think this topic is closer for my issue, but there's no answer here.

May be I misunderstood you? Thanks for your help.

[Ximera]

I returned to this question and added new route on my VBR Server: 10.22.229.0/25 via Proxy Appliance. After that static mapping for VM on second Production Network works perfectly.
Of course, such route is allowed for VBR Server only, not for core-routers or switches managing the whole infrastructure.

So I've come to conclusion that it's impossible to set up static mapped IP for VM on "not main" production network.