Database Load Strategy Selection for CPU Cost Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional database management systems (DBMS) face inefficiencies in optimizing CPU and memory resources, particularly during real-time analytics, as they are typically dedicated to transactional workloads rather than analytical workloads, leading to increased resource usage and decreased performance.
Innovation Solution
A computer-implemented method that determines the optimal load strategy for offloading data from a source DBMS to a target DBMS by calculating CPU costs and workload savings, selecting the best strategy to reduce CPU costs, and automatically removing unused tables to optimize memory and CPU usage, thereby enhancing workload performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is loaded into a target DBMS for accelerated analytical query execution, then workload performance is improved, but CPU cost for loading data increases
Solution Approach 1:
The system dynamically changes the parameter of data loading by evaluating multiple load strategies (full load, incremental load, partition-based load) and selecting the optimal one based on current workload characteristics, data size, and performance requirements, thereby resolving the contradiction between loading data for performance and CPU cost
Solution Approach 2:
The patent implements a dynamic decision-making process that continuously evaluates workload patterns, query performance metrics, and resource availability to adaptively determine when and how to load data into the target DBMS, transforming a static loading approach into a dynamic one that responds to changing conditions
2Speed
If tables are retained in the target DBMS for potential future queries, then query performance is improved, but memory usage increases
Solution Approach 1:
The system applies partial action by loading only the necessary subset of tables into the target DBMS based on workload analysis, rather than loading all tables. This resolves the contradiction by loading enough data to satisfy current analytical needs without unnecessarily consuming memory
Solution Approach 2:
The patent implements feedback mechanisms that monitor query performance, memory usage, and workload patterns to continuously evaluate whether retained tables are providing value. This feedback loop enables dynamic table removal decisions that balance query performance needs against memory consumption
Data Source
AI summary
Loading a set of data into a target database management system (DBMS) from a source DBMS to increase workload performance and decrease resource usage is provided. Using a size and a change rate of tables eligible to be offloaded to the target DBMS, a total central processor unit (CPU) cost corresponding to loading the set of data into to the target DBMS and workload cost savings corresponding to executing one or more workloads on the target DBMS is determined based on a set of one or more load strategies. The total CPU cost is compared with the workload cost savings of each load strategy. An optimal load strategy is selected based on comparing the total CPU cost with the workload cost savings of each load strategy. The set of data is automatically loaded into to the target DBMS from the source DBMS in accordance with the selected optimal load strategy.