High Availability
The Control Panel uses Raft-based clustering to ensure the fleet management system remains operational even if individual nodes go down. Settings → High availability (global admins) shows the current state of the cluster.
Cluster Status
The HA page displays:
- Node list — all nodes with their role (leader or follower) and status
- Leader indicator — which node is currently handling operations
- Replication status — whether all nodes are in sync
- Cluster health — overall health based on node availability
Node Statuses
| Status | Meaning |
|---|---|
| Active | Node is running and participating in the cluster |
| Syncing | Node is catching up on state replication |
| Unreachable | Node cannot be contacted |
Rolling Upgrades
When a platform upgrade is available, the HA system supports coordinated rolling upgrades to maintain availability throughout the process:
- Followers are upgraded one at a time
- Each node is taken out of the cluster, upgraded, and rejoined
- After all followers are upgraded, leadership transfers to an upgraded node
- The original leader is upgraded and rejoins
During a rolling upgrade, the cluster operates with reduced capacity but remains available. No downtime is expected.
Manual Failover
You can trigger a manual leadership transfer from Settings → High availability while this node is the leader and the cluster has at least two nodes:
- Click Trigger failover
- Confirm; Raft elects the new leader among the healthy followers
This is useful for planned maintenance or rebalancing.
Next Steps
- Fleet Management — manage your fleet
- Managed Hosting — provision servers with high-availability oversight