Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Elastic Cluster Management adds a topology-focused subpage to the server dashboard for Azure Database for PostgreSQL elastic clusters (Citus). Use it to inspect nodes, review table and shard layout, and generate management SQL that you can open in a query editor and review before running.
Open elastic cluster management
- Open the server dashboard for an Azure Database for PostgreSQL elastic cluster.
- Connect the dashboard if needed.
- On the dashboard toolbar, select Elastic Cluster Management.
The page opens inside the dashboard and starts with cluster-wide topology details. The top area includes Connection endpoints, which lists the cluster host and purpose-specific ports for coordinator access (5432 and 6432) and load-balanced access across cluster nodes (7432 and 8432).
Before executing generated DDL or cluster-wide management SQL in a query editor, use a coordinator connection: port 5432, or port 6432 for the coordinator's PgBouncer endpoint. Don't use the load-balanced ports 7432 or 8432 for those operations, because they can route the connection to a worker. Verify the query editor's active connection before running the SQL. See Elastic cluster concepts and Elastic cluster limitations.
Below that, the page organizes management data into four sections:
- Node management shows coordinator and worker nodes, shard counts, storage usage, and CPU when metrics are available.
- Table management lists distributed, reference, schema-sharded, and local tables.
- Shard management lists shard placements with search, sorting, and paging.
- Colocation groups summarizes tables that share compatible distribution metadata.
Use Refresh to reload the visible topology data.
Manage tables
Open Table management to browse tables by schema. The section is server-paginated, so changing schemas or pages reloads only the rows you need.
Selecting a table opens a details drawer. The drawer shows the table classification and topology details, then generates reviewable SQL for actions that apply to that table type, such as distributing a table, creating a reference table, registering a local table with Citus metadata, or undistributing a table or schema.
Elastic Cluster Management doesn't run those commands automatically. Instead, each action keeps the SQL in a read-only review surface and lets you open it in a connected query editor so you can inspect or modify it before execution.
Analyze shards
Open Shard management to search, sort, and page through shard placements. When you select a shard, a drawer opens and automatically runs a read-only shard analysis for that placement.
The analysis drawer combines several views:
- A shard summary for the selected placement.
- Tenant distribution when the shard belongs to a distributed table.
- A recommendation section that explains skew or placement findings.
- Generated SQL for actions such as splitting a shard or isolating a tenant.
When a recommendation is based on skew analysis, the drawer can also send the recommendation details to Copilot for explanation. As with table actions, generated SQL stays reviewable and opens in an editor instead of running automatically.