Query Plan Cost Estimation Through Runtime Statistics Feedback

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing query optimization techniques in database management systems (DBMS) face inaccuracies in estimating data statistics for query plans, leading to sub-optimal plan selection due to errors in estimating data statistics, especially for internal nodes, which can propagate and amplify estimation errors, resulting in inefficient resource usage.

Innovation Solution

Implementing a system and method for runtime statistics feedback that collects and applies accurate data statistics from previously executed query plans to improve cost-based query optimization by using a dictionary to map and propagate statistics across query trees, ensuring more precise cost estimation for alternative query plans.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If cost-based query optimization uses estimated statistics for query plan selection, then query optimization speed is improved, but measurement precision of data statistics deteriorates

Engineering Contradiction:
Improvequery optimization speedVSAvoiddata statistics estimation accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where actual execution statistics from previously executed query plans are collected and fed back to the query optimizer. This feedback loop allows the system to use accurate historical execution data to improve future cost estimates, resolving the contradiction between using fast estimates and achieving measurement precision.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary actions by collecting and storing execution statistics during query plan execution. These pre-collected statistics are then available for future cost-based optimization decisions, eliminating the need to wait for actual execution to gather accurate data while maintaining optimization speed.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If accurate runtime statistics are collected and applied from previous query executions, then measurement precision of cost estimation is improved, but device complexity increases

Engineering Contradiction:
Improvecost estimation accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system implements self-service by automatically collecting, storing, and applying execution statistics without requiring external intervention. The query optimizer autonomously uses the accumulated runtime statistics to improve cost estimates, reducing the need for manual configuration or external data sources while maintaining high measurement precision.

Inventive Principle:
Principle #25Self-service

3Productivity

If query plan cache stores compiled query plans for reuse, then productivity is improved, but reliability of optimal plan selection deteriorates due to estimation errors

Engineering Contradiction:
Improvequery execution efficiencyVSAvoidquery plan optimality
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent uses feedback from actual query plan executions to update and refine the statistics used for cost-based optimization. This continuous improvement loop ensures that cached query plans are selected based on increasingly accurate historical data, maintaining reliability while preserving the productivity benefits of query plan caching.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4296866B1Runtime statistics feedback for query plan cost estimation
Publication Date: 2025.08.06 SAP SE
  • EP4296866B1 patent drawingFigure 1
  • EP4296866B1 patent drawingFigure 2
  • EP4296866B1 patent drawingFigure 3

AI summary

A computer implemented method can execute a first query plan for a query, obtain statistics for internal nodes of a first query tree representing the first query plan, receive a second query tree representing a second query plan for the query, search for a matching internal node of the first query tree for a selected internal node of the second query tree, and responsive to finding the matching internal node of the first query tree, apply the statistics for the matching internal node of the first query tree to the selected internal node of the second query tree for estimating cost of the second query plan during query optimization of the query. Related systems and software for implementing the method are also disclosed.