Dynamic I/O Policy Control for Storage Performance Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data storage systems lack efficient methods to dynamically control and optimize performance according to specific performance goals, often requiring time-consuming reprogramming and not accommodating changing environments.
Innovation Solution
A method and system for creating and enforcing a performance policy that groups I/O requests into classes, monitors performance, sets goals, and adjusts processing to meet those goals, using a policy builder module to define and enforce I/O classes, performance metrics, and control methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data storage systems use fixed request processing numbers programmed in advance, then performance goals can be maintained, but the system cannot adapt to changing environments and uses
Solution Approach 1:
The patent implements dynamic performance control by transitioning from fixed programmed request processing to a flexible policy-based system. The system continuously monitors performance metrics and dynamically adjusts I/O request processing numbers based on current system conditions and policy definitions, enabling adaptation to changing environments while maintaining manageable complexity through automated control loops.
Solution Approach 2:
The system changes operational parameters dynamically by adjusting the number of I/O requests processed based on performance goals and current system state. Instead of fixed programmed values, the system modifies processing parameters in real-time based on monitored performance data, allowing the storage system to adapt to varying workloads and environmental conditions.
2Productivity
If data storage systems reprogram with revised request numbers to meet performance goals, then performance can be optimized, but the process is very time consuming
Solution Approach 1:
The patent applies preliminary action by pre-defining performance policies and goals before performance issues arise. The system has performance goals and policies configured in advance, allowing it to respond quickly to performance deviations by simply activating and adjusting these pre-prepared policies rather than performing time-consuming reprogramming operations when needed.
Solution Approach 2:
The system performs self-service by automatically monitoring its own performance, evaluating whether goals are met, and adjusting I/O request processing numbers without external intervention. This automated self-regulation eliminates the time-consuming manual reprogramming process, as the system independently optimizes its performance by modifying its own operational parameters based on monitored data.
3Ease of operation
If host systems control I/O requests directly, then performance can be optimized, but additional software must be installed on each host
Solution Approach 1:
The patent extracts performance control functionality from the host systems and consolidates it within the data storage system itself. By moving the performance monitoring and control logic into the storage system, the patent eliminates the need for additional host software while maintaining optimized I/O request processing, thereby improving ease of operation without adding host complexity.
Solution Approach 2:
The data storage system acts as an intermediary between host systems and disk drives, centralizing performance control functions. The storage system monitors performance metrics and adjusts I/O request processing numbers internally, serving as a mediator that optimizes performance without requiring host software modifications or direct host control, thus simplifying the overall system architecture.
Data Source
AI summary
Described are techniques for controlling performance of a data storage system. As a first part of an interactive dialogue, a policy is selected including one or more I/O classes each representing a logical grouping of I/O requests received at the data storage system. As a second part of the interactive dialogue, performance of said data storage system is monitored for each I/O class included in said policy without enforcing said policy. As a third part of the interactive dialogue, at least one performance goal associated with one of said I/O classes is set. The policy is run as a fourth part of the interactive dialogue causing enforcement of the policy on the data storage system in accordance with the at least one performance goal.