Dynamic Storage Array Selection via IO Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for selecting a primary storage array in a cluster based solely on latency between the host and storage arrays experience performance degradation due to high utilization, as they do not account for network latencies and utilization loads effectively.
Innovation Solution
A method and system that dynamically measure input/output servicing metrics and network latency across multiple storage arrays, establish thresholds, and dynamically select or switch the primary storage array based on these metrics to optimize performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If storage array selection is based solely on latency between host and storage arrays, then network response time is minimized, but performance degradation occurs due to high utilization not being accounted for
Solution Approach 1:
The patent changes the selection parameters from only latency to a composite set including latency, IO servicing metrics, and utilization load. This allows the system to dynamically adjust storage array selection based on multiple performance dimensions, resolving the contradiction by considering both network response time and storage performance simultaneously.
Solution Approach 2:
The system implements continuous monitoring of IO servicing metrics and utilization loads, using this feedback to dynamically adjust storage array selection. This feedback mechanism ensures that the system adapts to changing conditions, maintaining both low latency and high performance by selecting arrays that are currently least utilized.
2Productivity
If a single primary storage array is selected, then data access efficiency is improved, but performance degradation occurs when the selected array experiences high utilization
Solution Approach 1:
The patent makes the primary storage array selection dynamic rather than static. The system continuously monitors utilization metrics and can switch the primary storage array based on current conditions. This dynamic approach maintains data access efficiency by always selecting the most appropriate array while preventing performance degradation through adaptive reselection.
Solution Approach 2:
The system automatically monitors its own performance metrics and makes self-directed decisions about storage array selection without external intervention. This self-service capability allows the system to maintain optimal performance by independently adjusting its configuration based on real-time utilization data.
3Reliability
If multiple storage arrays are configured to process IO requests, then system capacity and redundancy are increased, but complexity of managing and selecting the primary array increases
Solution Approach 1:
The system automatically manages the complexity of multi-array configuration through self-directed monitoring and selection. By implementing automated metrics collection and decision-making algorithms, the system handles the complexity of managing multiple arrays without requiring manual intervention, thus maintaining high system capacity while minimizing management burden.
Data Source
AI summary
One or more aspects of the present disclosure relate to dynamically selecting a storage array and corresponding input/output (IO) paths between a host and the storage array. In embodiments, a virtual storage volume (VSV) can be established for a host entity using one or more storage device portions from a plurality of storage arrays. In addition, IO servicing metric parameters can be dynamically measured. For example, the servicing metric parameters can define metrics corresponding to the VSV's assigned ports on each storage array or network latency between the host and each storage array. Further, a primary storage array from the plurality of storage arrays can be selected based on the IO servicing metrics.