How to connect AI DBA Workbench to Citus  for intelligent monitoring

Most PostgreSQL monitoring tools have no concept of a Citus cluster - they see a coordinator and worker nodes as separate, unrelated servers. The pgEdge AI DBA Workbench treats them as what they are: a distributed system that needs to be watched as a whole.

This video walks through the setup and what you get out of it: adding your Citus coordinator and worker nodes as a single cluster, then watching the collector start building a picture of your system.

Across all your nodes, the Workbench tracks replication state, vacuum activity, query performance, WAL throughput, and connection health using 34 built-in probes and three-tier anomaly detection that catches the kinds of issues that don't show up as a single spiking metric until something's already wrong.

Ellie, the AI assistant built into the Workbench, is what kicks in when an alert fires. She's not just surfacing what changed - she reads the alert in the context of your full cluster state and walks you through a specific resolution: not a generic suggestion, but rather a query plan to inspect, a configuration to tune, or an index to add.

The Workbench connects over a standard Postgres connection, with no agents to install on your nodes and no changes to how your cluster runs today.

Open source under the PostgreSQL License. Built by the creator of pgAdmin.

Download: https://www.pgedge.com/download/ai-dba-workbench

GitHub: https://github.com/pgEdge/ai-dba-workbench