AutoTiering Controller for Multi-Tier All-Flash Data Center
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage systems using hard disk drives (HDDs) are inefficient when transitioning to flash storage, as they fail to optimize data placement across different storage tiers, leading to suboptimal access times and increased costs due to redundant caching and migration overhead.
Innovation Solution
The implementation of an AutoTiering system that uses an I/O Filter, Daemon, and Controller to monitor and migrate virtual machine data across multiple SSD tiers based on performance predictions and resource utilization, optimizing storage allocation and reducing migration costs by considering historical and predicted workloads, tier specializations, and service level agreements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is stored exclusively in the data center using flash storage, then storage efficiency is improved, but access time increases due to the reduced benefit of caching
Solution Approach 1:
The system segments storage into multiple tiers (hot, warm, cold) based on data access patterns. Frequently accessed data is placed in higher-performance tiers while less frequently accessed data is placed in lower-performance tiers, optimizing the balance between access time and storage efficiency.
Solution Approach 2:
The system performs preliminary actions by predicting future data access patterns and proactively placing data in optimal storage tiers before access occurs. This reduces actual access time by ensuring data is already in the correct location when needed.
2Loss of time
If redundant copies of data are stored in both cache and data center using flash storage, then access time is improved, but cost increases
Solution Approach 1:
The system applies local quality by providing different storage qualities to different data based on access patterns. Hot data receives high-performance storage while cold data uses lower-performance storage, eliminating the need for universal redundant caching and reducing overall cost.
3Ease of operation
If conventional storage devices are used without considering device specializations, then ease of operation is maintained, but productivity decreases due to suboptimal tier choices
Solution Approach 1:
The system implements self-service by automatically monitoring data access patterns, predicting future patterns, and making intelligent tier placement decisions without manual intervention. This maintains ease of operation while significantly improving productivity through optimized storage tier selection.
Solution Approach 2:
The system uses feedback mechanisms to continuously monitor actual data access patterns and compare them with predictions. This feedback is used to refine prediction models and adjust tier placements, improving productivity while requiring no user action.
4Quantity of substance
If data is frequently migrated across storage tiers, then storage efficiency is improved, but migration overhead increases
Solution Approach 1:
The system performs preliminary actions by predicting future data access patterns and proactively migrating data to appropriate tiers before access occurs. This reduces the frequency of reactive migrations and lowers overall migration overhead while maintaining storage efficiency.
Data Source
AI summary
A system for performing auto-tiering is disclosed. The system may include a plurality of storage devices offering a plurality of resources and organized into storage tiers. The storage devices may store data for virtual machines. A receiver may receive I/O commands and performance data for the virtual machines. A transmitter may transmit responses to the I/O commands. An auto-tiering controller may select storage tiers to store the data for the virtual machines and may migrate data between storage tiers responsive to the performance data. The selection of the storage tiers may optimize the performance of all virtual machines across all storage tiers, factoring the change in performance of the virtual machines and a migration cost to migrate data between storage tiers.