NSP Kubernetes implementation
Environment
A small-profile VM called the NSP deployer host holds the required NSP container runtime environment for NSP cluster instantiation. The required NSP software instances are deployed in Kubernetes pods to the cluster members, which are called NSP cluster VMs.
The NSP container environment uses only the storage local to each NSP cluster member; no cluster-wide network storage is required. The NSP service pods are deployed only to specific cluster members to ensure that each service has access to fixed portions of the local storage.
NSP deployment supports various levels of system redundancy; see Chapter 8, System redundancy and fault tolerance, the NSP Planning Guide, and the NSP Installation and Upgrade Guide for information.
Deployment profiles
Based on the deployment scope and required level of fault tolerance, you specify one of the following profiles during NSP cluster deployment, based on the response to your Nokia Platform Sizing Request:
-
node-labels-standard-3nodes.yml—standard deployment of three VMs
-
node-labels-standard-4nodes.yml—standard deployment of four VMs
As required to accommodate system or network growth, you can later add instances of functions such as MDM to an NSP cluster that has sufficient resources. See the NSP Installation and Upgrade Guide for information.
NSP cluster
The NSP cluster, a containerized set of Kubernetes hosts, is the central component of any NSP deployment. The cluster hosts the NSP Platform functions, and a significant number of NSP framework and application functions.
An NSP cluster may consist of only one member, or three or more that each host a portion of the installed NSP software. Each NSP cluster member is a designated Kubernetes worker node on which NSP services run. One or more members are also designated control nodes.
NSP functions hosted by NSP cluster members
The following installation options are deployable only in an NSP cluster: