Orchestrator at GitHub

This post introduces Orchestrator as GitHub's solution for managing MySQL failovers and replication topologies. It details Orchestrator's approach to failure detection by observing replication clusters rather than just individual masters, its failover decision-making process considering replication rules and server configurations, and its anti-flapping mechanisms. The post also elaborates on topology management for complex, multi-datacenter setups with intermediate masters and its chatops integration for visibility and operations.