Dynamic Load Balancing Across Storage Media Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems lack holistic load balancing methodologies that can dynamically distribute commands across multiple storage media devices based on their specific access rates, typically optimizing within a single drive rather than across the entire system.
Innovation Solution
A storage system and method that assigns weights to addressable storage regions based on fast or slow access rates, using a storage controller to identify and distribute loads only to weighted fast access regions, thereby optimizing load distribution across multiple storage media devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If load balancing is applied at macro level to networks outside storage systems, then network performance is improved, but storage system performance remains suboptimal because existing techniques only optimize within single drives
Solution Approach 1:
The patent divides the storage system into multiple storage domains, each with specific characteristics (fast access regions, slow access regions). Load balancing decisions are made locally based on domain-specific properties rather than treating the entire system uniformly. This allows optimization at the storage domain level while maintaining overall system performance.
Solution Approach 2:
The storage system is segmented into multiple storage domains with different access characteristics. Each domain can be independently managed and optimized. This segmentation enables the system to treat fast and slow access regions differently, improving overall throughput by routing loads appropriately across domains.
2Productivity
If all storage regions are used for load distribution, then capacity utilization is maximized, but performance degrades due to inclusion of slow access regions
Solution Approach 1:
The system dynamically adjusts which storage domains are available for load distribution based on current performance requirements. The availability of storage domains can change over time, allowing the system to optimize for speed when needed while still providing access to slower regions when appropriate. This dynamic approach resolves the contradiction between speed and capacity utilization.
3Productivity
If storage systems view individual drives in isolation with NCQ, then single drive optimization is achieved, but holistic system performance is lost
Solution Approach 1:
The patent merges the load balancing functionality across multiple storage domains into a unified system. Instead of each drive operating independently with NCQ, the system combines resources from multiple domains and applies centralized load balancing logic that considers system-wide conditions. This merging enables holistic optimization while managing complexity through a coordinated control approach.
Data Source
AI summary
A storage controller controlling a plurality of storage media devices receives one or more commands from a queue representing a load, identifies one or more weighted fast access storage regions within addressable storage regions across the plurality of storage media devices having a fast access rate, distributes load by utilizing only the weighted fast access storage regions within the addressable storage regions across the plurality of storage media devices having the fast access rate.