File System Management via Predictive Data Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for managing hybrid data storage systems with local and remote resources fail to optimally allocate storage resources, leading to long access times and potential data loss when network connectivity is unavailable, as they do not effectively anticipate the need to redefine data allocation for local fast memory access and manage remote storage efficiently.
Innovation Solution
A method that periodically recalculates the addressing of digital files based on metadata, access history, and local file structure, using a reference law that anticipates user usage patterns to control data transfer between local and remote memory, minimizing cache failures and bandwidth consumption, and employs deep learning algorithms to predict file usage and optimize storage allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is stored on remote cloud resources to increase storage capacity, then storage capacity is improved, but data access time deteriorates due to network dependency
Solution Approach 1:
The system performs preliminary actions by predicting which files will be needed based on user behavior patterns, application context, and access history. Files are pre-transferred from remote cloud storage to local storage before they are actually needed, so that when users need them, the files are already available locally, eliminating network wait time while maintaining the ability to store large amounts of data remotely
Solution Approach 2:
The system dynamically adjusts the distribution of data between local and remote storage based on changing conditions including user behavior patterns, application context, network availability, and storage usage. The prediction model continuously learns and adapts to user preferences, modifying which files are kept locally versus remotely, allowing the system to optimize the balance between storage capacity and access speed in real-time
2Loss of time
If files are kept on local storage to improve access speed, then data access time is improved, but storage capacity deteriorates due to limited local space
Solution Approach 1:
The system applies local quality by selectively placing only the most frequently accessed or predicted-to-be-accessed files on local storage, while keeping less frequently used files on remote cloud storage. This creates a differentiated storage strategy where hot files reside locally for fast access and cold files reside remotely for space efficiency, optimizing both access speed and storage capacity utilization
3Quantity of substance
If data is transferred between local and remote storage to optimize space, then storage capacity utilization is improved, but bandwidth consumption deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-transferring files to local storage based on predictions of future usage, rather than waiting for actual access requests. This reduces the frequency and volume of data transfers over the network, as files are moved proactively based on learned user patterns rather than reactively in response to access attempts, thereby reducing overall bandwidth consumption
Solution Approach 2:
The system uses feedback from user behavior patterns, application context, and access history to continuously refine its predictions about which files will be needed. This feedback loop allows the system to make increasingly accurate predictions, transferring only the necessary files to local storage and avoiding unnecessary data transfers, thus optimizing storage capacity utilization while minimizing bandwidth consumption
4Device complexity
If simple file management rules are used to allocate storage, then device complexity is reduced, but productivity deteriorates due to suboptimal resource allocation
Solution Approach 1:
The system applies self-service by automatically performing file allocation, transfer, and management decisions without requiring user intervention. The prediction model autonomously analyzes user behavior patterns, application context, and storage conditions to determine which files should be on local versus remote storage, handling the complexity of resource allocation internally while presenting a simple interface to users
Data Source
AI summary
A method for managing the file system of a computer terminal having a local memory, connected to a network comprising at least one remote storage device, comprising steps for periodically calculating the local or remote addressing of a digital file to be recorded, read or modified based on a periodically recalculated law, wherein said step of calculating the addressing is based on metadata present in said digital file to be processed, the access history of said file to be processed and the local tree structure of the files, said law is determined from a reference law, recalculated in the background, by processes based on a user's usage, to control the copying of a remote file to the local memory before access to said file is requested, and to erase local files for which the said law determines a usage indicator below a threshold value.