databaseinsights-get-advanced-aggregated-query-stats

About

The databaseinsights-get-advanced-aggregated-query-stats tool fetches aggregated performance metrics for queries executed on AlloyDB instances with Advanced Query Insights enabled. It ranks queries by total execution time (sum(execution_time) desc) to help identify top resource-consuming queries.

Compatible Sources

{{< compatible-sources >}}

Requirements

IAM Permissions

• databaseinsights.queryStats.fetch on the target location/project, which is granted by:
  • Database Insights Viewer (roles/databaseinsights.viewer)
  • Monitoring Viewer (roles/monitoring.viewer)

Parameters

parent (String, Required)

Project and location in the format projects/{project_id}/locations/{location}.

full_resource_name (String, Required)

The full resource identifier for the database instance (e.g., //alloydb.googleapis.com/projects/{project_id}/locations/{location}/clusters/{cluster_id}/instances/{instance_id}).

start_time (String, Optional)

Beginning of the time interval in RFC3339 format.

end_time (String, Optional)

End of the time interval in RFC3339 format.

database (String, Optional)

Filter stats to a specific database.

username (String, Optional)

Filter stats to a specific database user.

query_id (String, Optional)

Fetch stats for a single specific query hash.

page_size (Integer, Optional)

Maximum number of query stats to return (Default: 20).

page_token (String, Optional)

Token for fetching subsequent pages of results.

Example

YAML Configuration:

kind: tool
name: get_aggregated_query_stats
type: databaseinsights-get-advanced-aggregated-query-stats
source: database-insights-source

Sample CLI Invocation:

./toolbox invoke get_advanced_aggregated_query_stats \
  --prebuilt alloydb-postgres-observability \
  '{"parent":"projects/PROJECT_ID/locations/REGION","full_resource_name":"//alloydb.googleapis.com/projects/PROJECT_ID/locations/REGION/clusters/CLUSTER_ID/instances/INSTANCE_ID","page_size":10}'

Output Format

Returns a structured JSON object containing results (an array of row values) and metadata (field schema descriptors):

{
  "results": [
    [
      "1106633582131931382",
      "postgres",
      40324,
      20162,
      2,
      0,
      "SELECT BATCH_ID, ... FROM information_schema.schemata ..."
    ]
  ],
  "metadata": {
    "fields": [
      {"name": "query_id", "type": "STRING"},
      {"name": "database", "type": "STRING"},
      {"name": "sum(execution_time)", "type": "DOUBLE"},
      {"name": "avg(execution_time)", "type": "DOUBLE"},
      {"name": "sum(count)", "type": "DOUBLE"},
      {"name": "sum(wait_time)", "type": "DOUBLE"},
      {"name": "min(normalized_query_text)", "type": "STRING"}
    ]
  }
}

Reference

Additional Resources

• OneMCP Reference: get_advanced_aggregated_query_stats
• Use Database Insights with Model Context Protocol (MCP)