Automated Storage Tiering for Virtual Machine File Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In hyperconverged cloud storage systems, file system objects are often placed on a single tier of storage regardless of performance characteristics, leading to inefficient allocation of resources and increased costs due to disparities in performance and capacity, resulting in either excessive or insufficient resource allocation.
Innovation Solution
An automated tiering method that collects and analyzes performance data across storage devices to assign and move storage objects across different tiers based on their capabilities, optimizing placement for virtual machines within the same or different clusters to ensure the most economical and performant choice.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If file system objects are placed on a single tier of storage regardless of performance characteristics, then device complexity is reduced and ease of operation is improved, but resource allocation efficiency deteriorates and costs increase
Solution Approach 1:
The system implements dynamic storage tiering where file system objects are automatically moved between different storage tiers based on their performance requirements and workload characteristics. The automated tiering system continuously monitors performance data and adjusts object placement in real-time, transitioning from static single-tier placement to dynamic multi-tier placement that adapts to changing conditions.
Solution Approach 2:
The system changes the storage tier parameter for each file system object based on analyzed performance data and capabilities. By monitoring performance metrics and comparing them against storage device capabilities, the system automatically adjusts which tier each object resides on, optimizing the match between performance requirements and storage characteristics.
2Productivity
If automated tiering with performance analysis is implemented, then resource allocation efficiency is improved and costs are reduced, but device complexity increases
Solution Approach 1:
The storage system performs self-service through automated monitoring and analysis of performance data. The system autonomously collects performance metrics, analyzes them against storage capabilities, and makes placement decisions without requiring manual intervention. This self-managing capability handles the complexity internally while presenting a simplified interface to users.
Solution Approach 2:
The system implements continuous feedback loops where performance data is collected from storage operations, analyzed to determine optimal placement, and used to automatically adjust object tiering. This closed-loop feedback mechanism ensures that the system continuously optimizes resource allocation based on actual performance characteristics and changing workload conditions.
3Ease of operation
If storage objects are placed without analyzing performance data, then ease of operation is maintained, but performance optimization is lost and costs increase
Solution Approach 1:
The system performs preliminary analysis of performance data and storage capabilities before making placement decisions. By pre-analyzing performance characteristics and comparing them against storage tier capabilities, the system proactively determines optimal placement locations, ensuring performance requirements are met before objects are actually placed or moved.
Data Source
AI summary
In an example, a method of allocating storage objects of a virtual machine across storage devices in a computing system includes: collecting capabilities of the storage devices in the computing system; assigning a tier of a plurality of tiers to each of the storage devices in the computing system; collecting performance data for the storage devices across a period of time; analyzing the performance data and the capabilities to determine a placement of storage objects of the virtual machine across the plurality of tiers; and moving the storage objects to respective ones of the storage devices based on the assigned tiers.