Storage System Power Budget Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage systems, particularly flash storage, experience power usage spikes and surges due to operations like write and erasure, leading to inefficiencies and performance impacts when traditional power management methods like throttling are applied universally.
Innovation Solution
A method that monitors power usage at multiple levels of storage granularity, assigning non-uniform power budgets to storage units and adjusting sub-unit budgets dynamically based on usage, service levels, and policies to manage power distribution effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If power management approaches throttle erasure operations to reduce power consumption spikes, then power usage is reduced, but system performance deteriorates
Solution Approach 1:
The patent segments the storage system into multiple storage units with different power budgets, allowing granular control of power consumption at the storage unit level rather than system-wide throttling. This enables selective power management that maintains performance for critical operations while reducing power for non-critical ones.
Solution Approach 2:
The patent implements dynamic power budget allocation that adjusts power limits based on usage patterns, service level agreements, and operational priorities. This dynamic approach allows the system to provide full power during critical operations while limiting power during non-critical operations, resolving the contradiction between consistent performance and power reduction.
2Ease of operation
If uniform power budget is assigned to all storage units, then management is simplified, but power efficiency deteriorates due to inability to account for varying usage patterns
Solution Approach 1:
The patent assigns different power budgets to different storage units based on their specific usage patterns, service level agreements, and operational priorities. This local differentiation allows each storage unit to receive appropriate power allocation rather than a one-size-fits-all approach, improving overall power efficiency while maintaining manageable complexity through automated policies.
3Device complexity
If power budget is not adjusted at sub-unit level, then control granularity is reduced, but power management precision deteriorates
Solution Approach 1:
The patent further segments each storage unit into storage sub-units and implements power budget adjustment at this finer granularity level. This multi-level segmentation enables precise control of power consumption for specific sub-units within storage units, allowing the system to achieve high power management precision without overwhelming complexity through hierarchical management.
Solution Approach 2:
The patent adds another dimension to power management by implementing hierarchical control at both storage unit level and storage sub-unit level. This multi-dimensional approach allows simultaneous coarse-grained policy management and fine-grained execution control, achieving both manageability and precision.
Data Source
AI summary
A method includes monitoring power usage for a storage system that includes a set storage units at a first level of storage granularity and a set of storage sub-units at a second level of storage granularity, wherein the second level of storage granularity is finer than the first level of storage granularity. The method further includes assigning a non-uniform power budget to the set of storage units and adjusting a power budget for the storage sub-units according to the non-uniform power budget assigned to the storage units. A corresponding computer program product and computer system are also disclosed herein.