PLB VLB Defragmentation Amortization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern log structured storage systems face inefficiencies in garbage collection due to the need for large empty chunks to store user data, which involves user data relocation and creates overhead during the defragmentation process, particularly in managing physical layer blocks (PLBs) and virtual layer updates.
Innovation Solution
A method is implemented to identify and copy data from a maximum number of blocks to a new block, where the blocks are organized based on their utilized space, with the goal of minimizing the amount of space currently used, ensuring the sum of space utilized by the selected blocks equals the size of the new block, thereby optimizing the defragmentation process and reducing overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional garbage collection is used to create large empty chunks for storing user data, then storage capacity is improved, but user data relocation overhead and virtual layer update overhead increase
Solution Approach 1:
The patent segments the storage system into physical layer blocks (PLBs) and virtual layer blocks (VLBs), allowing independent management of each layer. PLBs are defragmented and consolidated at the physical layer without requiring virtual layer updates, thereby reducing overhead while maintaining storage capacity.
Solution Approach 2:
The patent extracts the defragmentation operation from the virtual layer and applies it directly at the physical layer. By working directly with PLBs instead of relocating user data through the virtual layer, the system eliminates unnecessary data movement and virtual layer updates.
2Productivity
If traditional defragmentation processes are used, then storage efficiency is improved, but the number of redirections and system overhead increase
Solution Approach 1:
The patent divides blocks into buckets based on their utilization characteristics (e.g., partially full blocks, mostly full blocks). This segmentation allows targeted defragmentation strategies for different block types, improving efficiency while reducing unnecessary redirections.
Solution Approach 2:
The patent performs preliminary organization of PLBs into buckets before defragmentation. By pre-categorizing blocks based on their space utilization, the system prepares the data structure to minimize redirections during the actual defragmentation process.
Data Source
AI summary
A method, computer program product, and computer system for identifying, by a computing device, a plurality of blocks. A maximum number of blocks of the plurality of blocks capable of being copied to a new block may be identified. Data from the maximum number of blocks of the plurality of blocks may be copied to the new block.