Use the Unused Indexes health check in Mini DBA to review PostgreSQL configuration, metadata, and monitoring signals before they become active alerts.
This health check looks for index or statistics conditions that can make query plans slower, less predictable, or harder to maintain.
Plan quality problems often appear as intermittent performance incidents. This check helps you find the design issue while there is time to tune it deliberately.
Review the affected objects and workload before changing indexes. Create or rebuild only where the workload justifies it, remove unused indexes carefully, and update statistics or maintenance routines so the condition does not return.
Keep before-and-after evidence such as execution plans, runtime, reads, writes, and index usage so the change can be judged by workload impact rather than preference.
Prioritize it by blast radius. Fix it quickly when it affects a production workload, a shared server, or a customer-facing service; otherwise place it in the normal improvement backlog with an owner.
Prevent repeat failures by adding the finding to the normal performance review cycle. Query and index checks are most useful when developers and DBAs share the same evidence and agree on the expected workload pattern.
Add this page to the operational runbook for PostgreSQL environments where the check repeatedly fails. Record local thresholds, approved exceptions, escalation contacts, and the expected review interval so the next administrator can act quickly without rediscovering the same context.