Features

A comprehensive toolkit for MySQL high availability and replication management.

Topology Discovery & Visualization

Automatically discovers MySQL replication topologies by crawling instances. Visualize your entire replication tree through the web UI with real-time status updates, lag monitoring, and interactive topology manipulation.

Automated Master Failover

GTID-based failover detects master failures and promotes the best candidate replica automatically. Configurable promotion rules, anti-flapping, and acknowledgment workflows ensure safe, predictable recovery.

ProxySQL Native Integration

Built-in hooks keep ProxySQL server groups synchronized with topology changes. When a failover occurs, orchestrator updates ProxySQL automatically so application traffic is rerouted with minimal disruption.

Prometheus Metrics & Kubernetes Health

Exposes a /metrics endpoint with replication lag, discovery cycle, recovery event, and backend health metrics. Kubernetes-ready liveness and readiness probes for container-native deployments.

API v2 with Structured Responses

A modern JSON API with consistent request/response envelopes, proper HTTP status codes, and structured error reporting. Manage topology operations, run recoveries, and query instance state programmatically.

MySQL & PostgreSQL Support

Full support for both MySQL and PostgreSQL replication topologies. PostgreSQL streaming replication is supported with automated discovery via pg_stat_replication, failure detection, and failover using pg_promote(). A provider abstraction layer makes it straightforward to add new database engines.

High Availability via Raft Consensus

Run multiple orchestrator nodes in a Raft cluster for leader election and state replication. If the active node fails, a standby takes over seamlessly — no single point of failure for your HA manager.

Additional Capabilities

  • Topology refactoring: Move replicas between masters, repoint replication, and restructure topologies through the UI or API.
  • Downtime management: Mark instances as in-maintenance to prevent automated failover during planned work.
  • Hooks and scripts: Execute custom scripts on topology events — pre/post failover, discovery, and more.
  • Pseudo-GTID support: Failover for legacy topologies without native GTID, using orchestrator's pseudo-GTID injection.
  • Cluster aliasing: Assign human-readable names to clusters for easier management and monitoring.
  • Access control: Role-based access through HTTP authentication, with read-only and read-write modes.
  • Key-value store publishing: Publish master discovery to Consul or ZooKeeper for service discovery integration.
  • CLI interface: Full-featured command-line tool for scripting and automation alongside the HTTP API.