The network requirements between active/standby NSP clusters depend on the network size (number of NEs and configured services) and the rate of service provisioning activities. The peak bandwidth requirement between redundant servers is 50 Mbps, with a sustained bandwidth of 25 Mbps. Round-trip network latency between the pair of redundant servers must be limited to 100 ms.

In a multi-node deployment of an NSP cluster, it is recommended that the nodes within the cluster have 1Gbps of Ethernet connectivity with less than 1 ms round-trip latency.

It is recommended that the link between the NSP deployer host and the NSP cluster has 1Gbps of Ethernet connectivity with less than 1 ms round-trip latency.

Note: This section applies to NSP deployments on customer provided k8s.

It is recommended that the link between the image registry to the k8s cluster has 1Gbps of Ethernet connectivity with less than 5 ms round-trip latency. The image registry is where the NSP docker images and NSP helm charts are hosted for customer provided k8s deployments.