Multi-level Non-volatile Cache Selective Store Mechanism
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-level non-volatile caches face limitations in effectively increasing storage capacity due to data replication across different storage levels, leading to inefficiencies in data management and access times.
Innovation Solution
A multi-level cache system with a selective store mechanism, where a higher storage level determines whether to store data, allowing the lower level to evict or maintain data based on an indicator, thereby optimizing storage and access by avoiding redundant data replication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is replicated across multiple storage levels in a multi-level cache, then data access reliability is improved, but storage capacity efficiency deteriorates
Solution Approach 1:
The patent extracts the data replication function selectively, applying it only to data that benefits from redundancy while leaving other data as single copies. The selective store indicator mechanism allows the system to take out the replication operation from the universal cache write path and apply it only when needed, thereby improving storage capacity efficiency while maintaining reliability for critical data.
Solution Approach 2:
Instead of fully replicating data across all storage levels for all data types, the patent applies partial replication only to specific data identified by the selective store indicator. This partial action approach maintains the reliability benefits of replication where needed while avoiding the storage capacity waste of universal replication.
2Reliability
If data is replicated across multiple storage levels, then data availability is improved, but data management complexity increases
Solution Approach 1:
The patent implements self-service by having the cache controller automatically manage replication decisions based on the selective store indicator without requiring complex external management. The system serves itself by using the indicator to automatically determine which data to replicate and which to store as single copies, reducing management complexity while maintaining data availability.
Solution Approach 2:
The selective store indicator is set in advance during data insertion, allowing the system to pre-determine replication needs before the write operation completes. This preliminary action simplifies subsequent data management by having replication decisions already made, reducing the complexity of real-time data management across multiple storage levels.
3Speed
If the lower storage level maintains all data copies, then data access speed is improved, but storage space utilization deteriorates
Solution Approach 1:
The patent segments the data set into two categories based on the selective store indicator: data that should be replicated for fast access and data that should be stored as single copies. This segmentation allows the lower storage level to maintain copies only for specific data segments, improving access speed for those segments while maintaining good storage space utilization overall.
Solution Approach 2:
Instead of applying uniform data retention policy across all data, the patent applies local quality by having the lower storage level maintain data copies only for specific locations in the storage hierarchy where it is beneficial. The selective store indicator enables local quality optimization where replication occurs only when and where it improves access speed without wasting storage space.
Data Source
AI summary
In embodiments, apparatuses, methods and storage media associated with a multi-level cache are described. A first storage level may receive an input/output (I/O) request from a second storage level of the multi-level cache, wherein the I/O request is associated with a data. The first storage level may further receive an indicator to indicate whether the data is stored or will be stored in the second storage level. The first storage level may determine whether to store the data in the first storage level based on the indicator. Other embodiments may be described and/or claimed.