Database Block Page Segmentation for Scan Area Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database management systems face challenges in simultaneously guaranteeing record reconfiguration and reducing the scan target area for efficient query processing, as previous techniques either fail to reduce the scan target area or do not ensure record reconfiguration.
Innovation Solution
A database configuration where each data block includes multiple data pages storing column values, allowing the database server to select a data block and specify a scan target data page, ensuring record reconfiguration and reducing the scan target area by storing all necessary column values within the same data block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If column values are stored in separate blocks for each column (PTL 2 approach), then the scan target area can be reduced, but record reconfiguration cannot be guaranteed
Solution Approach 1:
The database is segmented into data blocks, each containing multiple data pages. Each data page stores column values for a specific column, and all data pages belonging to the same data block are logically grouped together. This segmentation allows the system to read only necessary data pages (reducing scan target area) while ensuring that all column values for a record are contained within the same data block (guaranteeing reconfiguration).
Solution Approach 2:
The patent implements a nested structure where data pages are contained within data blocks, and columns are organized within data pages. This nested organization (data block → data page → column values) enables hierarchical access: the system can first identify the relevant data block, then specify the exact data page within that block, achieving both reduced scan area and guaranteed record integrity.
2Reliability
If all column values are stored in a single PAX page (PTL 1 approach), then record reconfiguration is guaranteed, but the scan target area cannot be reduced
Solution Approach 1:
Instead of storing all column values in a single PAX page, the patent segments the data into multiple data pages, each storing column values for a specific column. These data pages are grouped within data blocks, allowing the system to read only the necessary data pages for a given query while ensuring all required column values remain within the same data block for guaranteed reconfiguration.
Solution Approach 2:
The patent adds a hierarchical dimension to the data organization: data blocks contain multiple data pages, which in turn contain column values. This dimensional structure (block → page → column) enables the system to navigate and read only the specific pages needed for a query, reducing the scan target area while maintaining the integrity guarantee through the block-level organization.
3Quantity of substance
If data is stored in external storage apparatus, then cost per unit capacity is reduced, but processing speed for record reconfiguration decreases
Solution Approach 1:
The patent implements preliminary organization of data into data blocks with internal page-level structure before actual query processing. This pre-organization allows the database server to quickly identify and read only the necessary data pages from external storage, minimizing the I/O time and achieving fast record reconfiguration despite using cost-effective external storage.
Solution Approach 2:
The data block structure acts as an intermediary between the external storage apparatus and the query processing system. It provides a hierarchical indexing mechanism that enables efficient navigation and selective reading of data pages, bridging the speed gap between external storage access and in-memory processing requirements.
Data Source
AI summary
A database includes a plurality of data blocks. Each of the plurality of data blocks includes a plurality of data pages in which a plurality of column values recorded in one or more records corresponding to the data block are stored. Each of the plurality of data pages has two or more column values in one column corresponding to the data page stored therein. A database server selects a data block from the plurality of data blocks and specifies a data page to be scanned from the selected data block.