In the same way that server virtualization programmatically creates and manages virtual machines, NSX network virtualization programmatically creates and manages virtual networks.

With network virtualization, the functional equivalent of a network hypervisor reproduces the complete set of Layer 2 through Layer 7 networking services (for example, switching, routing, access control, firewalling, QoS) in software. As a result, these services can be programmatically assembled in any arbitrary combination to produce unique, isolated virtual networks in a matter of seconds.

NSX works by implementing three separate but integrated planes: management, control, and data. These planes are implemented as a set of processes, modules, and agents residing on two types of nodes: NSX Manager and transport nodes.
  • Every node hosts a management plane agent.
  • NSX Manager nodes host RESTful API services, management plane (MP) cluster daemons and central control plane (CCP) cluster daemons.
  • Transport nodes host local control plane (LCP) daemons and forwarding engines implementing the NSX data plane.

NSX supports a NSX Manager cluster with three nodes, which merges policy manager, management, and central control services on a cluster of nodes. Starting NSX release version 2.4, NSX controller elements resides in NSX Manager appliance. The convergence of management and control plane nodes reduces the number of virtual appliances that must be deployed and managed by the NSX administrator. NSX Manager clustering runs on top of a distributed data platform called Corfu and provides high availability of the user interface and the API .

The NSX Manager appliance is available in three different sizes for different deployment scenarios:
  • A small appliance for lab or proof-of-concept deployments.
  • A medium appliance for deployments up to 64 hosts.
  • A large appliance for customers who deploy to a large-scale environment.

See NSX Manager VM and Host Transport Node System Requirements and Configuration maximums tool.