Database Batch Update via Segmented Memory Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database access methods are limited by the efficiency of hard disk access, as the magnetic head needs to physically move to locate data, resulting in significant seek and rotation times, which hampers access rate.
Innovation Solution
A database batch update method that utilizes a memory storage apparatus with multiple memory types (non-volatile sequential access and volatile random access memories) to process data access commands, where sequential access is prioritized to improve access efficiency by separating data access types and mirroring memories for optimal sequential access rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data access commands are processed using conventional hard disk access methods, then data can be stored and retrieved, but the magnetic head must physically move to locate data, resulting in significant seek time and rotation time that hamper access rate
Solution Approach 1:
The patent segments data access commands into two types: sequential access commands and random access commands. Sequential access commands are processed by the sequential access memory system to maintain high access rates, while random access commands are handled by the random access memory system. This segmentation allows the system to optimize for sequential access performance without completely sacrificing random access capability.
Solution Approach 2:
The patent introduces a command classification mechanism as an intermediary between the data access request and the memory system. This intermediary classifies commands into sequential or random access types and routes them to appropriate memory systems, enabling the system to avoid the seek time and rotation time penalties of conventional hard disk access for sequential operations.
2Productivity
If sequential access method is adopted to improve hard disk access efficiency, then seek time and rotation time are decreased, but real-time data return commands may be interrupted
Solution Approach 1:
The patent segments the memory system into two independent subsystems: a sequential access memory system for processing sequential access commands, and a random access memory system for processing random access commands. This segmentation ensures that real-time random access commands are not interrupted by sequential access operations, as they are handled by a separate subsystem.
Solution Approach 2:
The command classification mechanism acts as an intermediary that separates sequential and random access commands before they reach the memory systems. This intermediary ensures that sequential access operations do not interrupt real-time random access data returns, as each type of command is routed to its dedicated memory subsystem.
3Productivity
If multiple memory types are used to separate data access types, then access efficiency is improved, but device complexity increases
Solution Approach 1:
The patent divides the memory storage apparatus into two distinct memory systems based on their access characteristics: sequential access memory and random access memory. Each memory system is optimized for its specific access type, allowing the system to maintain high access efficiency for both sequential and random operations without requiring a single complex unified memory system.
Solution Approach 2:
The command classification mechanism provides a universal interface that handles both sequential and random access commands. This intermediary layer abstracts the complexity of having multiple memory types, presenting a unified data access interface to users while internally routing commands to the appropriate memory subsystem.
Data Source
AI summary
A database batch update method is provided. The method includes: receiving a data access command which requires to access data from a first memory; determining whether the data access command is a first type or a second type command; storing the first type command in a second memory, and access the data from a third memory in response to the second type command; sequentially accessing the first memory according to the data access command stored in the second memory in an order of physical addresses of the first memory. The sequential access rate of the first memory is larger than the random access rate of the first memory, wherein the terms “sequential” and “random” are in connection with the physical addresses of the first memory. Furthermore, a data redo/undo log producing method and a memory storage apparatus are also provided.