Skip to main content

Registry & Control Plane High Availability

The service registry The service registry

The registry is the discovery hub that tracks controllers, control-planes, and clients. The Registry page gives you live visibility into your topology:

  • Instances — every active controller and control-plane, with its zone, version, node count, and uptime.
  • Leader election — see which instance currently holds leadership for scheduled work (snapshots, renewals).
  • Cleanup — remove a stale controller or control-plane that no longer reports in.

Multiple controllers share one MongoDB, elect a leader for singleton tasks, and standby nodes hydrate from registry snapshots — so the management plane keeps running through instance failures.

High availability at the edge

This page covers the management plane. To keep the data plane up when an edge box fails, pair two Elchi OS appliances so they share VIPs active/passive — see Edge HA Pairs. The edge pair keeps serving and failing over even while the control plane is down.