Accelerated Data Access via DRR Metadata Structures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Copy-on-write (COW) data access techniques lead to performance degradation due to traversal overhead, disk I/O amplification, and data memory amplification, as they require traversing multiple layers to find target data blocks, resulting in inefficient data access operations.
Innovation Solution
Implementing Data Relationship and Reference (DRR) data structures that utilize metadata from the COW layer hierarchy to accelerate data access by reusing in-memory copies of data blocks and reducing the scope of layer traversal, allowing direct access to existing copies without re-traversing the layer hierarchy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If COW technique is used to enable point-in-time logical view and data sharing, then data storage efficiency is improved through layering, but data access performance deteriorates due to traversal overhead
Solution Approach 1:
The patent pre-computes and stores metadata about data block locations across layers before access requests occur. When a data access request is made, the system uses this pre-stored metadata to directly locate data blocks without traversing the layer hierarchy, thereby resolving the contradiction between maintaining efficient layered storage and achieving fast data access.
2Quantity of substance
If COW technique divides dataset into layers for storage efficiency, then storage space is optimized, but data access operations become complex due to cross-layer dependencies
Solution Approach 1:
The patent introduces a metadata structure as an intermediary between the layered data storage and access operations. This metadata layer stores information about data block locations across different layers, allowing the system to maintain the storage efficiency benefits of layering while simplifying access operations by providing direct location information without requiring complex cross-layer traversal logic.
3Reliability
If COW technique performs traversal lookup across layers to find target data, then complete data access is ensured, but disk I/O operations are amplified
Solution Approach 1:
The patent pre-computes metadata containing information about data block locations across all layers before actual data access occurs. This preliminary action allows the system to ensure complete data access reliability by having all location information readily available, while simultaneously reducing disk I/O amplification by eliminating the need to traverse layers and read metadata from disk during actual data access operations.
4Measurement precision
If COW technique traverses multiple layers to locate target data blocks, then accurate data retrieval is achieved, but access time increases
Solution Approach 1:
The patent pre-computes and stores metadata about data block locations across layers before access requests occur. This preliminary computation ensures that when data access is needed, the system can directly locate the target data block using the pre-stored metadata without traversing multiple layers, thereby maintaining accurate data retrieval while significantly reducing access time.
Data Source
AI summary
Systems and methods are provided for implementing accelerated data access operations. For example, a method for accessing data includes loading an image into a memory, wherein the image comprises a plurality of image layers, and wherein each image layer references one or more data blocks. A request is received for a target data block from an active process operating on the image. Data structures associated with the target data bock are then accessed, wherein the data structures comprise metadata of a previous lookup path that was performed across one or more of the image layers of the image to access the target data block. The metadata of the data structures is utilized to access an existing in-memory copy of the target data block without having to traverse the previous lookup path, and the in-memory copy of target data block is returned to the active process.