Remote Database Mimic for Query Tuning Without Production Impact
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing query tuning methods impact production database performance and require significant human intervention, as they often rely on executing what-if analysis scenarios directly on the production database server, leading to resource constraints and inefficiencies.
Innovation Solution
A method that creates a remote mimic of the production database system, allowing query tuning analysis to be performed on this replica without affecting the production system, and applies recommendations automatically or with minimal human intervention, using cloud-based resources to offload the tuning process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If query tuning analysis is executed directly on the production database system, then tuning recommendations can be obtained, but production system performance is degraded and resource consumption increases
Solution Approach 1:
The patent creates a mimic database system that replicates the production database system's structure, data, and configuration. Query tuning analysis is executed on this copy rather than the production system, allowing accurate tuning recommendations to be obtained without impacting production performance. The mimic system is kept synchronized with the production system to ensure tuning results are applicable.
Solution Approach 2:
The mimic database system serves as an intermediary between the query tuning process and the production database system. It absorbs the computational burden of what-if analysis and tuning experiments, acting as a buffer that protects the production system from performance degradation while still enabling thorough query tuning.
2Measurement precision
If manual query tuning processes are used, then detailed analysis can be performed, but significant human intervention and time are required
Solution Approach 1:
The system implements automated query tuning where the mimic database system and tuning engine perform what-if analysis, evaluate multiple query plans, and generate recommendations without human intervention. The process autonomously compares different execution plans, assesses their performance characteristics, and presents optimized recommendations, eliminating the need for manual database administrator involvement in the analysis phase.
Solution Approach 2:
The system automatically evaluates query execution plans by simulating their performance on the mimic database system, compares results against defined criteria, and iteratively refines recommendations based on the feedback from these simulations. This automated feedback loop enables deep analysis without requiring human reviewers to examine each scenario.
3Adaptability or versatility
If multiple what-if analysis scenarios are executed on the production system, then comprehensive tuning options are explored, but system resources are overwhelmed
Solution Approach 1:
By creating a dedicated mimic database system copy, the patent enables multiple what-if analysis scenarios to be executed simultaneously without competing for production system resources. The mimic system absorbs all computational overhead from running multiple tuning scenarios, allowing comprehensive exploration of tuning options while production resources remain dedicated to actual workloads.
Solution Approach 2:
The patent segments the database system into two independent components: the production database system that handles actual workloads, and the mimic database system that handles tuning analysis. This segmentation allows resource-intensive what-if analysis scenarios to run on the mimic system without consuming production system resources, enabling comprehensive tuning scenario coverage.
Data Source
AI summary
Tuning a production database system through the use of a remote mimic. In response to receipt of a query tuning request against a database system, information about that system is obtained and a mimic of the system is set up in a remote system environment. The mimic aims to imitate the database system in all relevant ways with respect to the tuning request. A tuning analysis is then performed on this mimic system such that there is substantially no impact to operations of the original database system. Tuning results are then applied to the original database system. The entire process takes place with little or no human intervention.