Storage Controller Power Management for Data Center Energy Savings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers face significant power and cooling demands, and existing storage systems lack uniformity in power-saving features, making it difficult to implement energy-saving measures across all storage devices, especially those without proprietary power-down capabilities.
Innovation Solution
A storage controller determines inactive storage devices by lack of input/output requests and head load thresholds, powering them down and back up as needed, while performing maintenance during inactive periods to optimize energy usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If storage devices are kept powered on continuously to ensure immediate availability, then service availability is improved, but energy consumption increases
Solution Approach 1:
The system dynamically adjusts the power state of storage devices based on their activity status. Controllers monitor I/O requests and transition devices between active and powered-down states, optimizing the balance between availability and energy consumption. This dynamic approach allows the system to adapt to varying workloads rather than maintaining a static powered-on state.
Solution Approach 2:
The system implements periodic monitoring of I/O requests to storage devices and uses these periodic checks to determine when to power devices down or back up. By evaluating device activity at regular intervals and using head load count thresholds, the system creates a rhythmic pattern of power state transitions that reduces energy consumption while maintaining service availability.
2Use of energy by moving object
If storage devices are powered down to save energy, then energy consumption is reduced, but service availability deteriorates
Solution Approach 1:
The system performs preliminary actions by monitoring I/O request patterns and head load counts before actually powering down devices. Controllers track device activity metrics and predict when devices can be safely powered down, allowing proactive power management that prevents availability issues rather than reacting to them after they occur.
Solution Approach 2:
The system implements feedback mechanisms where controllers continuously monitor I/O requests and device activity status, then use this feedback to make informed decisions about power state transitions. The feedback loop ensures that devices are only powered down when it's safe to do so, and are quickly powered back up when needed, maintaining the balance between energy savings and service availability.
3Ease of operation
If power is continuously provided to all storage devices, then device availability is improved, but total cost of ownership increases
Solution Approach 1:
The system segments storage devices into different power management groups based on their activity patterns and criticality. Controllers can individually manage the power states of different devices rather than treating all devices uniformly, allowing less critical devices to be powered down while maintaining power to critical devices, thus reducing overall energy consumption and TCO.
Solution Approach 2:
The system changes operational parameters by transitioning storage devices between different power states (active, idle, powered-down) based on monitored conditions. By dynamically adjusting the power state parameter rather than maintaining a constant powered-on state, the system reduces energy consumption and associated costs while maintaining adequate device availability through controlled parameter changes.
4Use of energy by moving object
If storage devices are frequently powered up and down, then energy savings are achieved, but device lifespan may be affected
Solution Approach 1:
The system maintains copies of device state information and metadata about device activity patterns. By tracking head load counts and I/O request histories, the system creates a record of device usage that informs power management decisions, allowing it to optimize energy savings while being aware of the cumulative impact of power cycles on device lifespan.
Solution Approach 2:
The system performs preliminary assessment of device activity metrics and head load counts before initiating power down sequences. By evaluating whether devices have met minimum activity thresholds and are in safe states for power transition, the system prevents excessive or inappropriate power cycling that could harm device lifespan, while still achieving energy savings through controlled power management.
Data Source
AI summary
The present disclosure is directed to implementing power savings features on storage drives within a storage subsystem. A controller determines a drive is inactive and directs a power connector to prevent power from being provided to the drive. The controller may receive an input/output request for the inactive drive, direct the power connector to allow power to be provided, and provide the input/output request. When the controller receives an input/output request for the inactive drive, the controller sends a notification to the request's originator that the drive is unavailable and to retry after a fixed period of time. The controller performs maintenance on the drive when the drive is not inactive. The controller determines a maintenance time when the drive will be inactive and performs maintenance at an accelerated rate.