Storage Array Cooling Fan Control via I/O Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
External storage arrays face inefficiencies due to cooling fans operating at higher speeds than necessary, especially at lower altitudes, as they lack internal temperature measurement devices to adjust fan speed based on actual thermal needs.
Innovation Solution
A system that monitors input-output (I/O) metrics, such as read and write rates, to control cooling fan speed using a pattern-recognition model generated during a training phase, incorporating nonlinear nonparametric techniques and multiple input, multiple output (MIMO) controllers to correlate I/O metrics with storage array temperatures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cooling fan nominal speed is set high to ensure adequate thermal margin, then cooling reliability is improved, but power consumption increases and electrical efficiency deteriorates
Solution Approach 1:
The cooling fan speed is changed from a static high nominal speed to a dynamic speed that adjusts based on real-time I/O metrics. The system continuously monitors I/O metrics and adjusts fan speed accordingly, allowing the fan to operate at lower speeds during low-activity periods while maintaining adequate cooling during high-activity periods, thus resolving the contradiction between reliability and power consumption
Solution Approach 2:
The system changes the operating parameter of fan speed from a fixed high value to a variable value determined by I/O metric thresholds. By establishing multiple speed levels (first speed level for high I/O metrics, second speed level for low I/O metrics) and transitioning between them based on metric changes, the system optimizes the balance between cooling reliability and power consumption
2Use of energy by moving object
If cooling fan speed is reduced to lower power consumption, then electrical efficiency is improved, but cooling reliability deteriorates under high load conditions
Solution Approach 1:
The system implements a feedback mechanism where I/O metrics are continuously monitored and used to adjust fan speed. The controller receives feedback about system activity levels and automatically adjusts fan operation accordingly, ensuring that cooling reliability is maintained during high-load conditions while optimizing electrical efficiency during low-activity periods
Solution Approach 2:
The system performs preliminary classification of I/O metrics into different levels and pre-establishes corresponding fan speed levels. When high I/O metrics are detected, the system is already prepared to transition to higher fan speeds, ensuring cooling reliability is maintained without delay. This preliminary preparation allows the system to respond quickly to changing thermal conditions
3Measurement precision
If internal temperature measurement devices are added to enable temperature-based fan control, then cooling control precision is improved, but device complexity and manufacturing cost increase
Solution Approach 1:
The system introduces I/O metrics as an intermediary parameter that correlates with thermal conditions without requiring direct temperature measurement. By using I/O metrics (such as IOPS, throughput, or latency) as a proxy for thermal state, the system achieves effective fan control while avoiding the complexity and cost of internal temperature sensors
Solution Approach 2:
The system replaces the physical temperature measurement mechanism with a software-based I/O metric monitoring approach. Instead of using thermal sensors and temperature-based control logic, the system uses I/O performance metrics and algorithmic models to infer thermal conditions and control fan speed, thereby eliminating the need for additional hardware components
Data Source
AI summary
Some embodiments of the present invention provide a system that controls a cooling fan for a storage array. During operation, an input-output (I/O) metric of the storage array is monitored. Then, the cooling fan is controlled based on the I/O metric.