Memory-Mapped Data Expanse Files for Mainframe Database Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database management systems face limitations due to the need for caching data and managing locks, which restricts the size of database files that can be addressed, leading to issues as file sizes increase, and limit user applications to specific address ranges.
Innovation Solution
Implementing a data expanse viewer using memory-mapped files on a System Architecture Interface Layer (SAIL) based mainframe operating system, allowing database management systems to access larger database files by creating memory-mapped data expanse files, which provide a private view of the data without requiring caching or locking, enabling concurrent access by multiple user applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional database management systems use caching and locking mechanisms, then data access control is maintained, but the addressable range is limited and file size management becomes problematic
Solution Approach 1:
The patent introduces an operating system interface layer as an intermediary between the database management system and the file system. This layer provides memory-mapped file access that extends the addressable range beyond conventional limits while maintaining data access control through the operating system's memory management mechanisms, thus resolving the contradiction between reliability and adaptability
Solution Approach 2:
The patent transitions from conventional byte-addressing to memory-mapped file access, adding a new dimension of address space. By mapping files to memory regions, the system can access files larger than the traditional addressable range, effectively moving from a limited address space to an extended virtual memory space
2Speed
If database management systems cache data, then access speed is improved, but system complexity increases due to lock and test & set operations
Solution Approach 1:
The patent extracts the caching and locking functions from the database management system and relocates them to the operating system level. The memory-mapped file mechanism provides automatic caching through the operating system's virtual memory management, eliminating the need for application-level cache control and reducing system complexity
Solution Approach 2:
The memory-mapped file mechanism enables self-service data access where the operating system automatically manages the caching and memory management without requiring explicit lock or test & set operations from the database management system, thereby simplifying the control mechanism while maintaining access speed
3Device complexity
If the addressable range is limited in conventional systems, then simpler memory management is achieved, but the maximum size of database files that can be managed is restricted
Solution Approach 1:
The patent extends the addressable space by introducing memory-mapped files that utilize the operating system's virtual memory address space. This allows database files to exceed the conventional addressable range limits while the operating system's memory management mechanisms handle the complexity, thus increasing file size capability without proportionally increasing application-level complexity
Data Source
AI summary
Systems and methods for obtaining access to database files in a computing system. A method may include receiving a first call from a database management system requesting access to a database file. The method may further include transmitting a second call to an operating system interface requesting that a memory-mapped data expanse file be created. The method may also include receiving a first address representing the database file in response to successful mapping of the database file to the memory-mapped data expanse file located at the operating system interface.