Multi-Node Data Cache Segmentation for Remote Access Overhead Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-node computer systems, existing SSD cache technologies using set-associative mapping fail to identify affinity nodes, leading to increased remote access overheads and reduced cache efficiency, as data cannot be cached in nodes with which it has affinity, thereby decreasing data access efficiency.
Innovation Solution
Divide the cache area into thread and global cache areas, with associative mapping between the thread cache and disk array, and set-associative mapping between the global cache and disk array, allowing for detection of read frequencies and file sizes to determine optimal caching locations, ensuring files are cached in affinity nodes based on their access patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If set-associative mapping is adopted between SSD cache and magnetic disk, then data read and write efficiency is improved, but remote access overheads increase and cache efficiency decreases
Solution Approach 1:
The patent segments the cache area into multiple sub-areas, where each sub-area corresponds to a specific node in the multi-node system. This segmentation allows data to be cached in node-specific sub-areas, enabling the system to identify affinity nodes and reduce remote access overheads while maintaining cache efficiency.
2Loss of time
If data is cached in node-specific cache areas only, then remote access overheads are reduced, but cache utilization decreases
Solution Approach 1:
The patent applies local quality by creating node-specific sub-areas within the cache area, where each sub-area is optimized for local node access. This allows the system to maintain low remote access overheads for node-specific data while still enabling global cache utilization through the structured mapping approach.
Data Source
AI summary
A data cache method, device, and system in a multi-node system are provided. The method includes: dividing a cache area of a cache medium into multiple sub-areas, where each sub-area is corresponding to a node in the system; dividing each of the sub-areas into a thread cache area and a global cache area; when a process reads a file, detecting a read frequency of the file; when the read frequency of the file is greater than a first threshold and the size of the file does not exceed a second threshold, caching the file in the thread cache area; or when the read frequency of the file is greater than the first threshold and the size of the file exceeds the second threshold, caching the file in the global cache area. Thus overheads of remote access of a system are reduced, and I/O performance of the system is improved.