Dynamic Host Multipathing for Data Center Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current distributed storage systems face challenges in reducing power consumption while maintaining fault tolerance, particularly during off-peak load periods when underutilized components consume unnecessary energy.
Innovation Solution
The method involves using host multipathing software to identify redundant physical paths between host devices and storage arrays, and dynamically turning off underutilized paths during off-peak hours without violating redundancy policies, thereby reducing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple communication paths are maintained between host devices and storage arrays, then fault tolerance and reliability are improved, but power consumption increases
Solution Approach 1:
The system dynamically adjusts the state of communication paths based on load conditions. During peak load periods, all paths are kept active to ensure maximum fault tolerance. During off-peak periods, the system selectively turns off redundant paths while maintaining minimum required paths for fault tolerance, thus adapting power consumption to actual needs
Solution Approach 2:
The system changes the operational parameter of path availability from a static configuration to a dynamic state. By modifying the number of active paths based on workload conditions while respecting redundancy policies, the system optimizes the balance between reliability and energy consumption
2Reliability
If redundant communication paths are kept active, then fault tolerance is maintained, but energy waste increases during off-peak load periods
Solution Approach 1:
The system implements dynamic path management that responds to workload conditions. During off-peak periods, when storage arrays are underutilized, the system identifies and turns off redundant paths that would otherwise consume energy unnecessarily, while ensuring that sufficient paths remain active to maintain fault tolerance capabilities
Solution Approach 2:
The system extracts and removes redundant paths from the active configuration during off-peak periods. By identifying paths that exceed the minimum redundancy requirement and removing them from service, the system eliminates energy waste while preserving the essential fault tolerance function
Data Source
AI summary
A method, comprising: selecting an HBA entity that is currently turned on and which is part of a host device, the host device being coupled to a storage array, the host device being configured to read and write data to the storage array; identifying an identifier of the HBA entity; identifying, based on the identifier, a switch that is coupled to the HBA entity via a communications path; identifying one or more other communications paths that span between the host device and the switch; detecting whether a redundancy policy would be violated if the HBA entity were to be turned off, the redundancy policy including at least one rule that specifies a minimum number of active paths that must be present between the host device and the switch at any given time; turning off the HBA entity in response to detecting that the redundancy policy would not be violated.