Safe Write-Back Cache Selective Dirty Data Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing redundant storage systems face inefficiencies in data replication, particularly with write-back caching, where full mirroring of cache data leads to unnecessary replication of clean data, increasing costs and reducing cache performance.
Innovation Solution
Implementing a safe write-back cache that replicates only dirty data, using a dual-cache system where clean data is stored on one device and dirty data is duplicated across both devices, optimizing cache allocation and reducing redundancy for clean data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full mirroring of cache data is implemented, then data redundancy is ensured, but unnecessary replication of clean data increases costs and reduces cache performance
Solution Approach 1:
The patent extracts only the necessary portion of data (dirty data) for replication, leaving clean data unreplicated. The cache controller identifies and separates dirty data from clean data, replicating only the dirty data portion to the secondary cache device, thereby avoiding unnecessary replication overhead while maintaining data redundancy for modified data.
Solution Approach 2:
Instead of fully mirroring all cache data, the patent applies partial action by replicating only the necessary subset (dirty data). This partial replication approach provides sufficient redundancy for data that needs protection while avoiding the performance penalty of copying entire cache contents, including clean data that doesn't require replication.
2Reliability
If full mirroring of cache data is implemented, then data redundancy is ensured, but replication costs increase
Solution Approach 1:
The cache controller extracts and identifies only dirty data that requires replication, separating it from clean data. This extraction process enables selective replication where only the necessary data portion (dirty data) is copied to the secondary device, reducing the quantity of replicated data and associated costs while maintaining adequate redundancy.
Solution Approach 2:
The patent changes the replication parameter from full data replication to selective replication based on data state (clean vs. dirty). By monitoring data modification status and adjusting replication behavior accordingly, the system optimizes the quantity of replicated data, reducing overhead costs while maintaining necessary redundancy levels.
Data Source
AI summary
An embodiment of a cache apparatus may include a first cache memory, a second cache memory, and a cache controller communicatively coupled to the first cache memory and the second cache memory to allocate cache storage for clean data from one of either the first cache memory or the second cache memory, and allocate cache storage for dirty data from both the first cache memory and the second cache memory. Other embodiments are disclosed and claimed.