Storage Latency Database for I/O Command Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed storage systems face inefficiencies due to varying latency across storage devices, particularly in RAID arrays, which can lead to increased read/write latency during garbage collection and background tasks, affecting overall system performance.
Innovation Solution
A method that involves a latency database to monitor and manage latency by executing I/O commands differently based on latency thresholds, including default operations, read-by-recovery, and rebuilding the storage array when latency exceeds certain thresholds, thereby optimizing data access and storage operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the storage system executes I/O commands using default operations without latency monitoring, then the system structure remains simple, but the read/write latency increases during garbage collection and background tasks
Solution Approach 1:
The system performs preliminary actions by monitoring latency metrics and identifying when storage devices are experiencing elevated latency conditions before I/O commands are executed. The latency database pre-stores latency information for multiple storage devices, allowing the system to proactively select alternative storage devices before latency issues impact performance, rather than reacting after latency problems occur.
Solution Approach 2:
The patent introduces a latency database as an intermediary component that mediates between I/O command reception and storage device execution. This intermediary stores and provides latency information, enabling intelligent routing decisions. The mediator allows the system to balance between maintaining simple default operations and implementing latency-optimized operations by providing on-demand latency data without requiring complex real-time monitoring infrastructure.
2Productivity
If the storage system dynamically adjusts operations based on latency thresholds, then system performance is optimized, but the control mechanism complexity increases
Solution Approach 1:
The system applies parameter changes by dynamically selecting different I/O execution strategies based on latency threshold parameters. When latency for a storage device exceeds predefined thresholds, the system changes the operational parameter from default execution to alternative execution using different storage devices. This allows performance optimization through adaptive parameter adjustment without requiring complete redesign of the control architecture.
Solution Approach 2:
The patent implements dynamics by making the I/O command execution strategy adaptive rather than static. The system dynamically determines which storage device to use based on current latency conditions, switching between default and alternative execution paths. This dynamic behavior enables the system to respond to changing storage device performance conditions while maintaining a relatively simple control structure based on threshold comparisons.
3Speed
If the system uses alternative storage devices when latency exceeds thresholds, then data access efficiency is maintained, but the operation complexity increases
Solution Approach 1:
The system employs copying by maintaining multiple storage devices with redundant or alternative data copies. When the primary storage device experiences high latency, the system copies the I/O operation to an alternative storage device that has the same or equivalent data. This copying approach ensures data access efficiency is maintained by having backup execution paths without requiring complex data transformation or processing.
Data Source
AI summary
A method for use in a storage system, the method comprising: receiving an I/O command; identifying a latency of a first storage device that is associated with the I/O command; and executing the I/O command at least in part based on the latency, wherein executing the I/O command based on the latency includes: performing a first action when the latency is less than a first threshold, and performing a second action when the latency is greater than the first threshold, wherein identifying the latency includes retrieving the latency from a latency database, and wherein the first storage device is part of a storage array, the storage array including one or more second storage devices in addition to the first storage device.