Global Cache Management Table for Distributed Storage Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed storage systems, the lack of information about which data is stored in cache memory leads to inefficient utilization of cache resources across multiple storage apparatuses, resulting in reduced back-end bandwidth and responsiveness.
Innovation Solution
A storage system with multiple storage apparatuses, each equipped with a microprocessor, memory, and storage devices, where a shared area manages data distribution and cache allocation based on a global cache management table, allowing data to be dynamically allocated and reallocated across cache areas based on importance and availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a master node inquires of all nodes whether data exists in cache memory, then data access requests can be handled, but communication overhead and processing time increase significantly
Solution Approach 1:
The system pre-establishes a global cache management table that records which data is stored in which node's cache memory before access requests occur. This preliminary information registration eliminates the need for real-time inquiries to all nodes, allowing the master node to directly determine data locations based on pre-collected information.
Solution Approach 2:
A global cache management table serves as an intermediary data structure that mediates between the distributed cache memories of multiple nodes and the master node. Instead of direct communication between the master node and all cache nodes, the management table acts as a centralized information repository that provides quick lookup capabilities.
2Speed
If data is stored in local cache memory only, then access speed is improved, but overall cache utilization efficiency decreases
Solution Approach 1:
The system merges the distributed cache memories of multiple storage nodes into a unified global cache system. The global cache management table consolidates information about all cache resources, enabling the system to treat distributed caches as a single pooled resource that can be efficiently allocated across the entire storage system.
Solution Approach 2:
The system dynamically allocates and manages cache resources based on actual access patterns and availability. The global cache management table is updated in real-time to reflect changes in cache content and availability, allowing the system to adaptively optimize data placement and cache utilization across multiple nodes.
3Loss of energy
If cache memory is used to store frequently accessed data, then back-end bandwidth consumption is reduced, but cache memory space is limited
Solution Approach 1:
The system extends the cache capacity dimension by utilizing distributed cache memories across multiple storage nodes. Instead of being constrained by the cache size of a single node, the system aggregates cache resources from multiple nodes, effectively increasing the total available cache space while maintaining low back-end bandwidth consumption.
Data Source
AI summary
A first storage apparatus, on the basis of management information representing data existing in a virtual cache area, when a first data conforming to a read request exists in the virtual cache area, determines whether the first data is stored in a first cache area of the first storage apparatus. The first storage apparatus, when the first data is not stored in the first cache area, acquires the first data from a second cache area of a second storage apparatus, and transmits the first data to a host computer. The first storage apparatus, when the first data is associated with the first cache area on the basis of a degree of importance, notifies the second storage apparatus to that effect, and change an apparatus in charge of a cache of the first data of the management information from the second storage apparatus to the first storage apparatus. The first storage apparatus, when the first data is associated with the second cache area, notifies the second storage apparatus to that effect, and deletes the first data.