Log-Based Storage Management for Sequential Media
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Shingled magnetic recording (SMR) techniques and other sequentially written data storage devices face challenges in efficient data deletion and storage density, leading to increased data fragmentation and wasted space, as direct deletion is impractical and non-sequential writing is constrained by physical and logical limitations.
Innovation Solution
A log-based storage design that tracks data deletions by adding references to deleted data in subsequent records, allowing for compaction and consolidation of data, enabling efficient data management on sequentially written devices by transferring data to other storage devices when deletion thresholds are met, and using indexing for quick data retrieval and defragmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data deletion is performed on sequentially written media by updating records, then data deletion capability is maintained, but storage space does not increase proportionally with deletion level and data fragmentation increases
Solution Approach 1:
The patent segments the storage medium into multiple zones or regions, allowing deleted data in one zone to be compacted and reused while maintaining sequential write operations in other zones. This segmentation enables space reclamation without requiring complete sequential rewriting of the entire medium.
Solution Approach 2:
The patent introduces an intermediary indexing layer that maps logical data addresses to physical storage locations. This intermediary structure allows the system to track deleted data, perform logical deletions, and identify reusable space without disrupting the sequential write constraint, effectively decoupling deletion operations from physical space management.
2Ease of manufacture
If data is rewritten sequentially on sequentially written media, then write operations are simplified, but data fragmentation increases when deleted portions are replaced
Solution Approach 1:
The patent performs preliminary compaction operations to consolidate valid data and eliminate fragmentation before new write operations begin. By pre-processing the storage medium to create contiguous valid data regions, the system maintains sequential write simplicity while preventing fragmentation accumulation.
Solution Approach 2:
The patent dynamically adjusts write parameters such as block size, allocation unit, and compaction threshold based on the current fragmentation level and deletion pattern. This adaptive parameter adjustment optimizes the balance between sequential write efficiency and fragmentation control.
Data Source
AI summary
Techniques for implementing a data management scheme for optimizing data storage are described herein. A deletion quantity or other metric relating to deletions of data within a first storage zone are tracked. Upon detection that the tracked deletion metric meets certain criteria, the data within the first storage zone are moved to a second storage zone. A verification of the data to be moved is also performed, and if such verification indicates that at least a portion of the data is corrupted, routines repairing and/or restoring at least the corrupted portion are initiated.