Query Stats Detail

The Query Stats Detail dashboard focuses on a single query (text or normalized text) and shows how each compiled plan for that query performed over the selected interval.

Top: Query text

• The full SQL text (or normalized text) is shown at the top for context.
• Use the copy and download controls to move the text into SSMS or a local editor.
• Use the format button to view a formatted version of the query text

Plans summary table (Totals by plan)

• A compact table that lists each plan compiled for this query with totals and averages:
  • Execution count
  • Worker time (total)
  • Total time (total)
  • Averages (worker time / exec, total time / exec)
  • Rows returned
  • Memory grant (total)
• Use the table to quickly identify the highest-cost plans and their relative impact.

Totals section

• This section contains a time-series table of 5-minute aggregated samples for the selected time frame. Each row represents a 5-minute bucket and includes:
  • Sample time
  • Execution count
  • Logical reads
  • Logical writes
  • Memory
  • Physical reads
  • Rows
  • Total time
  • Worker time
  • Plan hash (clickable)
• Plan hash actions:
  • Click a plan hash to download the plan as a .sqlplan file.
  • Open the downloaded .sqlplan in SSMS to inspect the graphical plan and operator costs.
• Charts in Totals:
  • Execution count by plan
  • Memory by plan
  • Total time by plan
  • Worker time by plan
• Use these charts to compare plan behavior over time and to spot periods where a specific plan dominated resource usage.

Averages section

• Similar to Totals, but values are averaged per sample or per execution:
  • The table shows the same 5-minute samples with averaged metrics (per-exec averages where applicable).
  • Columns mirror the Totals table (execution count, reads, writes, memory, physical reads, rows, total time, worker time, plan hash).
• Charts in Averages:
  • Total time (avg) by plan
  • Worker time (avg) by plan
• Use averages to find plans with high per-execution cost even if execution counts are low.

Controls and investigation tips

• Use the dashboard time-range controls to focus on the interval of interest.
• Filter or sort the tables by plan hash, execution count, or cost metrics to prioritize investigation.
• Download and open .sqlplan files in SSMS to review operator costs, warnings, and estimated vs actual row counts.
• Compare Totals and Averages: high totals with low averages usually indicate frequent cheap executions; high averages suggest expensive single executions.