Network pools automatically assign static IP addresses to vMotion, vSAN, and NFS vmkernel ports so that you don't need to enter IP addresses manually when creating a VI workload domain or adding a host or cluster to a workload domain.

A network pool is a collection of a set of subnets within an L2 domain. Depending on the storage option you are using, it includes information about subnets reserved for the vMotion and vSAN or NFS networks that are required for adding a host to the Cloud Foundation inventory.
Table 1. Information Required for a Network Pool
Storage Being Used Required Networks in Network Pool
vSAN vMotion and vSAN
NFS vMotion and NFS
vSAN and NFS vMotion, vSAN, and NFS
VMFS on FC vMotion only or vMotion and NFS

The network pool also contains a range of IP addresses, called an inclusion range. IP addresses from the inclusion ranges are assigned to the vMotion and vSAN or NFS vmkernel ports on the host. The use of inclusion ranges allows you to limit the IP addresses that will be consumed from a given subnet. You can add more inclusion ranges in order to expand the use of the provided subnet.

A default network pool (named bringup-networkpool) is created during bring-up. This network pool is automatically associated with the management domain. Network information for this network pool is based on the deployment parameter sheet you provided during bring-up. This network pool contains vMotion and vSAN networks only - an NFS network is not supported in this network pool. If you have a single L2 domain in your environment for management workload domain vSAN and vMotion networks or if you want to expand the management domain by adding a host, you can expand this default network pool.

In order to create a workload domain with hosts in a different L2 domain than the management domain, or if you want to use external NFS or VMFS on FC storage, you must create a new network pool. A network pool can contain both vSAN and NFS networks.

For NSX for vSphere workload domains, all hosts in a cluster must be associated with the same network pool. For NSX-T workload domains, you can use the Cloud Foundation API to select hosts from different network pools, as long as those network pools have the same VLAN ID and MTU settings.

All hosts in a cluster must be associated with the same network pool. However, a workload domain can contain multiple clusters, each with its own network pool. You may want to have multiple clusters within a workload domain to provide separate fail over domains (i.e. a VM only fails over between hosts in a cluster). Multiple clusters also provide isolation for security reasons and are also useful for grouping servers of a particular type of configuration together. Multiple clusters can also be used to handle growth. Original servers used in the first cluster may get outdated at some point. Newer server models can then be added in a new cluster to the workload domain and workloads can be migrated at a leisurely pace.