The high availability solution is used to provide redundancy within a single site. NSP cluster deployment supports high availability of critical services through replica pods in a container environment. Specific pods are deployed with multiple replicas.

The high availability solution uses a few different approaches to managing a failure.

• Some services are deployed using multiple load-balanced instances where the load handled by any single instance can be picked up by the remaining instances should that single instance fail.

• Some services are deployed in a 3:1 model where there are three instances of the service running and one of them is active. If that active instance fails, then one of the remaining two standby instances will take over immediately.

• For services that are not replicated, additional resources are provided so that a new instance can always be instantiated whenever a failure occurs.

Note: In an enhanced/HA deployment, if node4 were to go down due to an ungraceful shutdown (such as a power outage), a switchover would be triggered.