Dynamic Index Mapping for Sequential I/O Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional disk arrays using RAID technology struggle to distribute user I/O operations evenly across all disks, leading to longer operation times for sequential I/O operations, as the mapping relationship between logical blocks and RAID stripes is fixed, preventing the use of added disks for parallel processing.
Innovation Solution
A method that involves receiving an access request for a logical block, determining the corresponding index entry from a plurality of index entries indicating the arrangement of logical blocks on a pool stripe, and processing the request using the target disk block, allowing for improved parallel processing of sequential I/O operations by utilizing index information to direct data access across multiple disks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed mapping relationship between logical blocks and RAID stripes is used, then the disk array structure is simple and stable, but sequential I/O operations cannot be distributed evenly across all disks, leading to longer operation times
Solution Approach 1:
The patent transforms the fixed mapping relationship into a dynamic one by introducing an index entry mechanism that can adaptively map logical block addresses to different pool stripe locations. When disks are added to the disk pool, the index entries are updated to reflect the new disk configuration, enabling sequential I/O operations to be distributed across all available disks including newly added ones, thereby improving I/O performance without requiring a completely fixed static structure
Solution Approach 2:
The patent introduces index entries as an intermediary layer between logical block addresses and physical disk locations. These index entries contain mapping information that enables the system to translate logical addresses to specific pool stripe locations dynamically. This intermediary mechanism allows the system to maintain a simple logical interface while achieving complex distributed I/O across multiple disks, resolving the contradiction between simplicity and performance
2Quantity of substance
If disks are added to the disk pool, then storage capacity increases, but the fixed mapping relationship prevents new disks from being utilized for parallel processing of sequential I/O operations
Solution Approach 1:
The index entry mapping structure is designed to be dynamic and adaptable to changes in disk pool configuration. When new disks are added, the system updates the index entries to incorporate the new disk locations into the mapping relationship. This allows sequential I/O operations to automatically utilize the expanded storage capacity through parallel processing across all disks including newly added ones, eliminating the limitation of fixed mappings
3Loss of time
If the mapping relationship is fixed at the RAID stripe level, then the disk array structure is stable, but operation time for sequential I/O increases due to inability to distribute operations across all disks
Solution Approach 1:
The index entry serves as a flexible intermediary that maintains stable logical block addressing while enabling dynamic physical distribution. The index entries can be updated to reflect new disk configurations without changing the logical block address space, thus maintaining stability at the logical level while achieving time reduction through improved physical distribution for parallel processing
Data Source
AI summary
Techniques for processing data involve receiving an access request for a logical block used for storing data, the access request including a logical block address for the logical block. Such techniques further involve determining, if it is determined that the logical block address is in a used address space, an index entry corresponding to the logical block address from index entries, the index entry indicating arrangements of logical blocks on a corresponding pool stripe, wherein the pool stripe includes disk blocks from disks in a disk pool, and data blocks form at least one data stripe for a disk array. Such techniques further involve determining a target block corresponding to the logical block based on the index entry. Such techniques further involve processing the access request by using the target block. Accordingly, the processing speed of sequential I/O operations is improved, time is saved, and user experience is improved.