Designing mappings for node interfaces - MediaLive

Designing mappings for node interfaces

This section is intended for the video engineer who is responsible for designing the MediaLive Anywhere workflows. In each cluster, you must design a collection of mappings. Each mapping connects the network interfaces on the nodes to the network that each interface uses. You must design these mappings after you have identified the networks and designed the clusters.

The mapping works as illustrated in the diagram that follows. In this example, you need to create a logical connection from the interface that handles inputs on the node to the network that handles input. You therefore create a mapping that consists of two pieces of information:

  • A logical interface name for the physical interface. For example, my-Inputs-Interface.

  • The ID of the network that handles inputs.

You will create the mappings in the cluster. Then in each node, you will create a second mapping that assigns each logical interface name to the appropriate node interface.

Network, cluster, and node diagram showing logical interface mappings and network ID connections.

  1. Assign IDs to the networks and the network logical interface names. Keep in mind that name fields are case-sensitive. Make a list of these IDs and names, and make sure that you use these exact names when you later create the networks and create the logical interface names.

  2. Design one mapping for all the network interfaces that handle a type of traffic. Don't create a mapping for each node.

    For example, create one mapping for the network interface that handles inputs. All the nodes will use the same mapping. The interface could theoretically be Eth1 on some nodes and Eth2 on other nodes. It's a logical mapping, so the number of the network interface isn't important.

  3. Make a list of the mappings for the cluster, one mapping for each node interface and each network that the cluster is attached to. This list captures the information in the blue boxes (one per cluster) in the diagram.

    For example:

    • Cluster A, Network ID=1234567, Logical Interface name=my-Inputs-Interface

    • Cluster A, Network ID=9876543, Logical Interface name=my-Outputs-Interface

    • Cluster B, Network ID=1234567, Logical Interface name=my-Inputs-Interface

    • Cluster B, Network ID=9876543, Logical Interface name= my-Outputs-Interface

    In this example, cluster A and cluster B share the same two networks. In your deployment, clusters might not share networks.

    You will use this information to complete the Interface mappings fields when you create each cluster.

  4. Make a list of the mappings for each node, one mapping for physical interface and each cluster. This list captures the information in the green box in the diagram. You create only one mapping for each physical interface because all the nodes in one cluster will be set up identically. For example:

    • Nodes in cluster A: Logical Interface name=my-Inputs-Interface, interface=Eth1

    • Nodes in cluster A: Logical Interface name=my-Outputs-Interface, interface=Eth2

    • Nodes in cluster B: Logical Interface name=my-Inputs-Interface, interface=Eth2

    • Nodes in cluster B: Logical Interface name=my-Outputs-Interface, interface=Eth3

    In this example, the nodes in cluster A use Eth1 and Eth2, but the nodes in cluster B use Eth2 and Eth3.

    You will use this information to complete the Node interface mapping fields when you create the nodes.