Database Idle-Time Page Compression for Space Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems face inefficiencies in space utilization due to the need for long-term locking of database segments during data compression and reorganization, which is impractical for high-performance online transaction processing applications requiring continuous access.
Innovation Solution
The method involves locking and compressing data pages during database idle time, recording the freed space, and allowing other pages to remain accessible, thereby eliminating the need for reorganization commands like reorg+rebuild, allowing compressed space to be used during runtime without locking the entire table.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data compression and reorganization is performed using reorg+rebuild command combination, then database space utilization is improved, but database access is locked for relatively long periods of time
Solution Approach 1:
The patent segments the database into multiple data pages and processes them individually rather than locking the entire database. The compression process locks only specific data pages (e.g., every nth page) while leaving other pages accessible, thereby reducing the overall lock duration and improving space utilization without prolonged access restrictions.
Solution Approach 2:
The patent implements periodic compression by selectively compressing every nth data page during idle time periods. This periodic approach allows the database to maintain normal access to uncompressed pages while gradually compressing data pages over time, resolving the contradiction between space optimization and continuous access requirements.
2Quantity of substance
If data compression and reorganization is performed continuously, then database space is reduced, but fast and continuous access to the database is disrupted
Solution Approach 1:
The patent dynamically adjusts the compression process by identifying and locking only specific data pages that require compression while leaving other pages accessible. This dynamic selective locking allows the database to maintain high productivity and continuous access speeds for uncompressed pages while still achieving space reduction through targeted compression of specific pages during idle time.
3Quantity of substance
If reorg+rebuild command combination is used for data compression, then storage space is optimized, but system cost increases due to operational complexity
Solution Approach 1:
The patent implements self-service compression by automatically identifying data pages suitable for compression and processing them during idle time periods without requiring manual reorg+rebuild operations. The system autonomously manages the compression process, selecting every nth data page for compression and updating the database structure accordingly, thereby reducing storage space efficiently while minimizing operational complexity and system costs.
Data Source
AI summary
Methods, systems, and computer program products are provided to manage a database system. The method includes locking during a database system idle time access by the database system to a data page of a data allocation unit, compressing during the database system idle time a data stored in the locked data page, and recording during the database system idle time an indication that the compressed and locked data page includes free storage space, wherein unlocked data pages of the data allocation unit are accessible by the database system during the compressing of the data stored in the locked data page. Thus, the data page may be compressed during idle time and the space freed therein may be used during a subsequent run time without the need for a reorganization of the data pages within the corresponding table (as in, for example, operation of a reorg+rebuild SQL command combination).