Azure Local SDN Stack error during Cluster Creation

Question

So I have a 2-node system with a network-switch for storage. I have configured on both nodes a net-intent (name is "allcombined") that combines compute, management and storage traffic in one interface. I now have 2 vNIC's on each node, one vManagement and one vSMB. These were created when assigning the netintent to a pNIC. According to my answers on my previous questions on the forums, the vSMB NIC gets an IP in the range off 10.71.1.x/24. This is an IP-address that should internally be handled by the SDN-stack withing Azure Local (Storage intent for Azure Local - Microsoft Q&A, How does SDN work in Azure Local? - Microsoft Q&A). The external staff said that I shouldn't worry about it and let it be because of the backend handling it. Now I get the following error:

When viewing the StorageClusterValidationRaport I see the following info:

This basically means that the two vNIC's that were automatically created using networkATC can't reach eachother. I was wondering if anyone knows a bit more about this because the external staff told me not to worry about this part of the configuration

Greetings Jaco

Answer 1

Hi ,

Thanks for reaching out to Microsoft Q&A.

The primary error message states: "Network interfaces a1-Node1.jaczazurelocal.local - vSMB(allcombinedPort1) and a1-Node2.jaczazurelocal.local - vSMB(allcombinedPort1) are on the same cluster network, yet address 10.71.1.87 is not reachable from 10.71.1.104 using UDP on port 3383." (And the reverse for the other node).

This clearly indicates a connectivity issue between the vSMB virtual NICs on your two nodes. Even though they are supposed to be on the "same cluster network," they cannot communicate. The fact that it's using UDP on port 3383 suggests it's likely related to SMB Direct (RDMA) or some internal cluster communication mechanism for storage.

The core issue is that while the 10.71.1.x/24 range might be "internally handled" for specific SDN functions, for the cluster to form and function correctly, these vSMB interfaces absolutely must be able to communicate with each other.

The error "address is not reachable using UDP on port 3383" for your vSMB network interfaces indicates a critical communication failure between your Azure Stack HCI nodes. Despite external advice, these internal SDN-managed IPs (10.71.1.x) must be able to communicate for cluster creation. This issue is likely caused by a firewall blocking UDP 3383, incorrect VLAN tagging, or a physical switch misconfiguration preventing traffic flow between the nodes on the ports associated with the "allcombined" intent. You must troubleshoot and resolve this network connectivity to proceed with cluster creation.

Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the reply was helpful. This will be benefitting other community members who face the same issue.