Adaptive Cache Partitioning for Multi-User Virtual Local Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Shared computing devices face challenges in storing user data due to security and privacy concerns, as well as physical storage size limitations, making it difficult for multiple users to utilize cloud storage effectively.
Innovation Solution
A self-adaptive cloud storage cache partitioning mechanism that provides virtual local storage (VLS) to each user, utilizing a Pre-boot Read-Write Proxy and Run-time Read-Write Proxy to manage data blocks at the platform level, ensuring near-native input/output performance and data privacy through a self-adaptive cache partitioning strategy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cloud storage is used for shared computing devices, then data security and privacy are improved, but access speed and latency are worsened
Solution Approach 1:
The patent segments cloud storage into user-specific partitions, where each user has a dedicated partition for their data. This segmentation allows for optimized access patterns while maintaining security boundaries, resolving the contradiction between secure cloud storage and fast access by organizing data in a way that enables efficient retrieval within partitioned structures.
Solution Approach 2:
The system performs preliminary actions by pre-loading and caching frequently accessed data blocks into local memory before they are actually needed. This proactive data preparation reduces access latency when users retrieve their data from cloud storage, as the most commonly accessed information is already positioned for rapid retrieval.
2Reliability
If cloud storage is used for shared computing devices, then data privacy is improved, but storage capacity is worsened
Solution Approach 1:
The patent implements a nested storage architecture where local storage partitions are embedded within the broader cloud storage structure. Each user has a virtual local storage (VLS) partition that is nested within their cloud storage space, creating a hierarchical organization that provides both local-like access speed and cloud-based capacity.
Solution Approach 2:
The system creates copies of frequently accessed data blocks and stores them in local cache partitions, while the original data remains in cloud storage. This copying mechanism allows users to access data locally with fast performance while the full storage capacity is maintained in the cloud, resolving the contradiction between privacy-protected cloud storage and available storage capacity.
3Speed
If virtual local storage is implemented for each user, then access performance is improved, but system complexity is worsened
Solution Approach 1:
The patent introduces a partition manager as an intermediary component that handles all operations between users and the partitioned storage system. This single intermediary manages partition creation, data block allocation, cache management, and access coordination, simplifying the overall system architecture despite the presence of multiple user partitions by centralizing control logic.
Solution Approach 2:
The partition manager is designed as a universal component that performs multiple functions: it manages partition boundaries, handles data block allocation and deallocation, coordinates cache operations, and enforces access policies across all user partitions. This multi-functionality reduces system complexity by consolidating what could be multiple separate management systems into a single versatile manager.
4Productivity
If cache partitioning is optimized for each user, then cache hit rate is improved, but computational overhead is worsened
Solution Approach 1:
The patent implements self-service mechanisms where the partition manager autonomously monitors cache performance metrics and automatically adjusts partition configurations to optimize cache hit rates. Instead of requiring external intervention or complex real-time computations, the system self-adjusts by tracking access patterns and redistributing cache space to users with higher access frequencies, reducing computational overhead while maintaining high cache efficiency.
Data Source
AI summary
An example shared computing device comprises a network interface to provide access to a first cloud storage including first user data associated with a first user and a second cloud storage including second user data associated with a second user, a local storage including a first partition with a first partition size, where the first partition stores a subset of the first cloud storage, and a second partition with a second partition size, where the second partition stores a subset of the second cloud storage, and a partition manager to cause a change to the first partition size based on a first cache rate of the first partition and a second cache rate of the second partition, where the cache rates are a percentage of the first cloud storage and the second cloud storage stored in the respective partition.