HTAP Database Execution Planning Across Row-Column Storage Modes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current database systems struggle to efficiently support hybrid transactional and analytical processing (HTAP) due to interference between transactional and analytical loads in hybrid row-column storage deployment, and high costs in row-column storage separation.
Innovation Solution
A database system that supports both hybrid row-column and row-column storage separation deployment modes, allowing flexible selection and enabling the computing system to formulate execution plans for each mode, utilizing memory-level, block-level, and column storage modules to optimize data storage and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If hybrid row-column storage deployment mode is used to support both transactional and analytical processing, then the database can handle HTAP workloads, but transactional and analytical loads interfere with each other
Solution Approach 1:
The patent segments the storage system into distinct row storage modules and column storage modules, allowing transactional operations to access row-stored data while analytical operations access column-stored data, thereby isolating the two types of workloads and eliminating mutual interference
Solution Approach 2:
The patent implements dynamic deployment modes that can switch between hybrid row-column storage and row-column storage separation based on workload requirements, enabling the system to adapt its storage configuration in real-time to optimize performance for different operational scenarios
2Reliability
If row-column storage separation deployment mode is used to eliminate load interference, then transactional and analytical processing are isolated, but system costs increase
Solution Approach 1:
The patent designs a unified storage system that can operate in multiple deployment modes (hybrid and separated), allowing the same infrastructure to serve both transactional and analytical workloads efficiently without requiring completely separate systems, thereby reducing overall system costs
Solution Approach 2:
The system dynamically selects between hybrid and separated storage deployment modes based on real-time workload characteristics, using separation only when necessary to eliminate interference, thereby minimizing the costs associated with maintaining separate storage paths while still achieving load isolation when needed
3Quantity of substance
If hybrid row-column storage deployment mode is used, then storage space efficiency is improved, but deployment flexibility is reduced
Solution Approach 1:
The patent implements a dynamic deployment architecture that can switch between hybrid row-column storage mode (for space efficiency) and row-column storage separation mode (for flexibility), allowing the system to adapt its deployment configuration based on specific operational requirements while maintaining both efficiency and flexibility capabilities
Data Source
AI summary
A computing system receives a deployment mode selection instruction for a target database, where the deployment mode selection instruction instructs the target database to use a hybrid row-column storage deployment mode or a row-column storage separation deployment mode in a storage system. The computing system determines, based on the deployment mode selection instruction, an execution plan corresponding to the target database, where the execution plan matches a deployment mode indicated by the deployment mode selection instruction. The computing system processes, in the storage system, a database task for the target database based on the execution plan.