Dynamic Storage Allocation in Tiered Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fat provisioning in tiered storage systems leads to low capacity utilization, increased seek times, and performance degradation due to pre-allocation of storage space, which results in wasted capacity and inefficient data distribution across different tiers.
Innovation Solution
A method that reserves storage capacity without initial allocation, allocating it on demand in response to write I/O operations, and migrates data based on access frequency and time, using a tiered hierarchy of storage devices such as SSDs, SAS, and SATA drives to optimize data placement in high-performance tiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If storage space is pre-allocated for users or applications, then capacity utilization is improved, but waste of unused space occurs
Solution Approach 1:
The patent implements dynamic provisioning where storage capacity is not statically allocated but dynamically assigned based on actual usage needs. The system monitors data placement patterns and automatically migrates data between tiers, allowing storage resources to adapt to changing demands rather than remaining fixed in pre-allocated spaces.
Solution Approach 2:
The system changes the provisioning parameter from static pre-allocation to dynamic on-demand allocation. By implementing thin provisioning with automatic tier migration based on access patterns, the system transforms storage management from a fixed parameter state to a variable one that optimizes capacity utilization while minimizing waste.
2Loss of time
If storage capacity is pre-allocated, then data can be immediately accessed, but seek times increase and performance degrades
Solution Approach 1:
The system performs preliminary actions by pre-positioning frequently accessed data in high-performance tiers before access requests occur. Through automated monitoring of access patterns, the system proactively migrates hot data to faster storage media, ensuring optimal access times without requiring full pre-allocation of capacity.
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor data access patterns and automatically adjust data placement accordingly. The system uses access frequency and timing information to make real-time decisions about data migration between tiers, creating a closed-loop control system that optimizes both access time and overall performance.
3Ease of manufacture
If data is scattered across logical volume regions, then initial write operations are simplified, but seek times increase when accessing different regions
Solution Approach 1:
The patent applies local quality by optimizing data placement based on access patterns rather than uniform distribution. Frequently accessed data is concentrated in high-performance tiers with optimized placement, while less accessed data resides in lower tiers. This creates locally optimized storage regions that balance write simplicity with read performance.
Data Source
AI summary
Systems, methods, and computer-readable media for handling I/O operations in a storage system are described herein. An example method includes assigning each of a plurality of storage devices to one of a plurality of tiers; imposing a hierarchy on the tiers; creating a logical volume by reserving a portion of a storage capacity for the logical volume without allocating the portion of the storage capacity to the logical volume; and assigning the logical volume to one of a plurality of volume priority categories. The method includes receiving a write I/O operation directed to a logical unit of the logical volume; and allocating physical storage space for the logical unit of the logical volume in response to the write I/O operation. The physical storage space is located in one or more storage devices. The method includes writing data associated with the write I/O operation to the one or more storage devices.