Query Analysis¶

Query Analysis is a standalone analysis entry point in the Database module. It supports cross-instance query aggregation. When you manage multiple database instances, you can identify system-level slow query bottlenecks from a global perspective without opening each instance's details page.

Capabilities¶

Unlike the Explorer, which analyzes queries on a single-instance basis, Query Analysis provides the following core capabilities:

• Cross-instance aggregation: When the same business SQL is executed on multiple instances, it is automatically aggregated into global cumulative data, so you do not need to troubleshoot instance by instance.
• Similar query consolidation: The system automatically identifies similar SQL statements with different parameter values (for example, WHERE id = 1 and WHERE id = 2) and consolidates them into a single query template for statistics, preventing the same statement from being scattered due to parameter differences.
• Global performance ranking: Supports sorting by total time, execution count, average time, and other dimensions to quickly identify global performance bottlenecks.

Page Layout¶

The Query Analysis page consists of a top action bar, left-side quick filters, a query list, and a details side panel.

Top Action Bar¶

Left-Side Quick Filters¶

Query List¶

The list shows all aggregated query templates for the current time range and database type. Each record represents one class of SQL (with parameters replaced by ?) and aggregates the execution data of that class across all selected instances.

List Fields¶

Query Details Side Panel¶

Click any row in the list to slide out the Query Details side panel on the right, which shows the global performance data of this query class. The side panel contains the following tabs:

Complete SQL¶

The top of the side panel displays the complete SQL template of this query class (with parameters replaced by ?), with one-click copy supported.

Performance Trend¶

Linked to the time range of the list page, the performance trend displays three core metrics of this query class as time series charts:

Each metric is displayed as an independent time series chart. You can select a time interval to filter the data or export the chart.

Instances¶

Shows the execution distribution of this query class across instances, helping you determine which instance the bottleneck is concentrated on.

• The instance name is clickable and opens the single-instance query details page for that instance in a new tab (reusing the existing Explorer details).
• You can filter by instance name or database address in the search box.

Users¶

Shows the execution distribution of this query class across database accounts.

• You can filter by user name in the search box.

Query Samples¶

Shows the actual sampled records of this query class. The fields vary by database type.

Example fields for MySQL:

Execution Plan¶

Click a SQL template in the query list to open the details side panel and go to Execution Plan. The execution plan describes the execution method chosen by the database optimizer for the current query. You can use it to determine whether the query has problems such as full table scans, improper index usage, or high JOIN costs.

Selecting an Execution Plan¶

The same query may use different execution plans on different database instances, and multiple plans may exist on the same instance. You can switch between them using the selector at the top of the page. The options consist of an instance name and a plan identifier, for example prod-sql-01 · 0x72AE91F3.

With SQL Server, for example, the plan identifier is displayed as Plan Hash. When the Plan Hash of the same query changes, it usually indicates that the database has adopted a different execution plan.

After you select a plan, the page shows the database instance, database name, and plan identifier associated with the plan.

Topology Graph¶

The topology graph shows the database execution steps according to the node hierarchy in the execution plan. The connections between nodes represent the dependencies between nodes, not the actual execution timeline of the query.

Each node shows the following information based on the data returned by the database:

• The execution operation name;
• The logical operation (shown only when the original plan contains this field and it differs from the execution operation);
• The table or index accessed;
• Rows: the number of rows the database optimizer estimates the node will output;
• Cost: the relative cost of the current node and its child nodes.

You can zoom in, zoom out, drag, or fit the entire topology graph to the view. Hover over a node to view additional information such as the execution operation, logical operation, accessed object, estimated rows read, filter conditions, JOIN conditions, or native database warnings. The page only shows fields that actually exist in the original plan.

Node List¶

The node list shows the node hierarchy of the execution plan with indentation, making it easy to quickly compare the accessed objects, estimated rows, and node costs across nodes. The list order expresses the hierarchical relationships between nodes, not the actual execution time order.

XML / JSON¶

The raw view shows the complete execution plan obtained by the collector. It supports formatting, searching, copying, and downloading.

Notes on Metric Calculation¶

The aggregated metrics in Query Analysis are calculated as follows: