Dynamic Storage Tiering Using Allocate-on-Write Snapshots
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing dynamic storage tiering systems face challenges in relocating data while it remains online and accessible, as they often require downtime or inefficient block size management across different storage tiers.
Innovation Solution
The method involves creating a point-in-time copy of a virtual volume with a storage hot-spot, copying the relevant segment from a first storage pool to a second pool, and reconfiguring the logical block address mapping to reference the copy in the new pool, allowing for dynamic relocation of data without downtime and optimizing block size for efficient tiering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is relocated between storage tiers while online, then data accessibility is maintained, but system complexity increases due to snapshot management and mapping reconfiguration
Solution Approach 1:
The patent creates a point-in-time copy (snapshot) of the virtual volume before relocating data. This copy serves as a reference during the migration process, allowing the system to maintain data accessibility while managing the complexity of tier relocation through a simplified copying mechanism rather than complex real-time synchronization protocols.
Solution Approach 2:
The snapshot acts as an intermediary structure that mediates between the source and destination storage tiers. By introducing this intermediate copy, the system can perform mapping reconfigurations and data transfers without directly coupling the complexity of tier management with online accessibility requirements, thus resolving the contradiction between reliability and complexity.
2Productivity
If traditional storage tiering methods are used, then block size management is simplified, but data relocation requires downtime and reduces productivity
Solution Approach 1:
The system performs preliminary actions by creating a snapshot of the virtual volume before initiating data relocation. This preparatory step captures the state of the volume at a specific point in time, enabling the subsequent migration to proceed without requiring the volume to be taken offline, thus eliminating downtime and maintaining productivity.
Solution Approach 2:
The patent implements dynamic storage tiering where data can be relocated between tiers while the system remains operational. The logical block address mapping is dynamically reconfigured during the migration process, allowing the system to adapt to changing storage requirements without interruption, thereby improving productivity while minimizing time loss.
3Power
If high-activity data is relocated to high-performance tiers, then system performance improves, but storage capacity in high-performance tiers is reduced
Solution Approach 1:
The patent applies local quality by identifying specific high-activity data segments (hot-spots) within the virtual volume and relocating only those segments to high-performance storage tiers. This selective approach ensures that system performance is improved for the most critical data while the majority of low-activity data remains in lower-performance tiers, thus preserving overall storage capacity in the high-performance tier.
Solution Approach 2:
The virtual volume is segmented into different data segments based on activity levels. The system identifies and isolates high-activity segments for relocation to high-performance tiers, while low-activity segments remain in lower-performance tiers. This segmentation allows the system to optimize performance for critical data without sacrificing overall storage capacity, as only portions of the volume are migrated.
4Manufacturing precision
If point-in-time snapshots are created during data migration, then data consistency is maintained, but additional storage resources are consumed
Solution Approach 1:
Instead of creating snapshots of entire storage volumes, the system applies partial action by creating snapshots only of the specific virtual volume segments that are being migrated. This reduces the storage resources consumed by snapshot metadata and copy-on-write structures while still maintaining data consistency for the relevant portions of the volume during the migration process.
Data Source
AI summary
Systems and methods for dynamic storage tiering using snapshot functionality are disclosed. A point-in-time copy of a virtual volume including a storage hot-spot is created; write operations directed to the virtual volume may be redirected to a point-in-time temporary virtual volume. The virtual volume segment, including the hot-spot, is copied from a first storage pool to a second storage pool. Finally, a logical block address mapping of the virtual volume is reconfigured to reference the virtual volume segment copy in the second storage pool. Upon deletion of the point-in-time copy of the virtual volume, the virtual volume segment copy in the second storage pool may be updated with data from the point-in-time temporary virtual volume.