Support leader election when running multiple replicas

**Describe the feature**
I'd like NTH v2 to use leader election when more than one replica is running.

**Is the feature request related to a problem?**
I'd like to have multiple replicas running to enable fast switching between the active replica and the standby without waiting for the scheduler.

**Describe alternatives you've considered**
n/a

The issue names no files, tests, or entry points. Start by tracing the existing replica and scheduler behavior in the Kubernetes integration; done means multiple replicas safely elect one active leader and support fast switching to a standby.

Written by the indexing model from the issue text.