Database Computing Node Memory Partitioning for Lock-Free Query Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database systems face limitations in processing speed due to hardware constraints, data storage methods, and operating system deadlocks, which hinder efficient execution of complex database functions.
Innovation Solution
A parallelized database system architecture with multiple sub-systems, including data input, storage and processing, query and response, administrative, and configuration, utilizing a combination of computing devices and nodes with independent processing core resources, optimized by an application-specific operating system to manage resources efficiently and prevent deadlocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple operating systems share computing resources, then resource utilization increases, but system reliability decreases due to potential deadlocks
Solution Approach 1:
The patent segments the operating system into multiple independent layers (kernel layer, system service layer, and application layer), where each layer operates independently with defined interfaces. This segmentation prevents deadlocks by isolating resource management functions in the kernel layer from user-space operations, allowing resource sharing without compromising system stability.
Solution Approach 2:
The patent introduces a file system abstraction layer as an intermediary between the kernel and user applications. This intermediary manages file access and resource allocation, enabling multiple operating systems to share resources safely by mediating their interactions and preventing conflicting operations that could lead to deadlocks.
2Speed
If traditional operating systems are used, then hardware compatibility is maintained, but processing speed decreases due to system calls and context switching
Solution Approach 1:
The patent merges the benefits of traditional operating systems with specialized database processing capabilities into a unified multi-layer architecture. The kernel layer maintains hardware compatibility through standard device drivers, while the system service layer provides optimized database operations, achieving both speed and compatibility.
Solution Approach 2:
The patent changes the operational parameters of the operating system by implementing asynchronous I/O operations and optimized memory management in the system service layer. These parameter changes enable faster data processing while the kernel layer maintains standard hardware interfaces for compatibility.
3Speed
If data is stored using conventional methods, then storage capacity is maximized, but access speed decreases due to sequential I/O operations
Solution Approach 1:
The patent transforms the traditional sequential I/O model into a parallel processing model by implementing multiple I/O channels and asynchronous operations. This dimensional change from sequential to parallel access enables faster data retrieval while maintaining storage capacity through efficient use of I/O resources.
Solution Approach 2:
The patent implements data pre-fetching and caching mechanisms in the system service layer, performing preliminary actions to load frequently accessed data into memory before it is needed. This reduces I/O wait time and speeds up data access without requiring additional storage capacity.
Data Source
AI summary
A computing node of a computing device of a computing device cluster of a store and compute sub-system of a database system includes main memory and processing core resources. The main memory is logically divided into a computing device memory section and a database memory section. The database memory section includes a first logical section for a first processing core resource, a second logical section for a second processing core resource, and a logical network communication section. The processing core resource retrieves first long-term storage data of a first division of a sub-segment of a segment of a segment group of a partition of a dataset and retrieves a set of query operations. The processing core resource executes the set of query operations on the first long-term storage data to produce a first division result and stores it.