High Availability

High availability (HA) for Canton nodes means running replicated instances that share a database so a passive replica can take over when the active instance fails. Use these howtos when you need resilient participant or synchronizer infrastructure.

Party multi-hosting improves resilience for a party, not for a node process. For node-level failover, use the replicated setups below.

Guides

Participant high availability — active/passive participant replicas, fail-over, and load balancer health checks
Synchronizer high availability — mediator replicas and horizontally scaled database sequencers

Requirements

Shared storage between replicas (PostgreSQL; Oracle where noted on the participant page)
Replication / sequencer HA settings as described in each howto

Console: replication.set_passive and synchronizers.connect_multi
Observability configuration for health endpoints used by load balancers