Redundant Liquid Cooling Loops for Storage Rack Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Information handling systems face risks of coolant leakage in liquid cooling solutions, leading to potential data loss and unavailability, necessitating a high availability solution that prevents whole rack shutdowns due to single point failures.
Innovation Solution
A system with redundant converged liquid loops and directional switching mechanisms, allowing for continuous cooling by switching to a secondary loop in case of failure, ensuring 100% cooling capacity for storage nodes without shutting down the entire rack.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single cooling loop is used for storage rack, then the device complexity is reduced, but the reliability deteriorates due to risk of coolant leakage causing whole rack shutdown
Solution Approach 1:
The cooling system is divided into multiple independent cooling loops, each capable of providing full cooling capacity to all storage nodes. This segmentation allows isolation of failures to individual loops while maintaining overall system reliability, as each loop is designed to handle 100% of the cooling load independently.
Solution Approach 2:
Switching valves are pre-configured and positioned to enable rapid transition between cooling loops before a failure propagates. The system maintains standby cooling capacity through pre-established redundant loops, allowing immediate failover without shutdown when a leakage or failure is detected.
2Reliability
If redundant cooling loops are implemented, then the reliability is improved by preventing whole rack shutdown, but the device complexity increases
Solution Approach 1:
Each cooling loop is designed with universal capability to serve all storage nodes independently. The switching valves and cooling distribution units are configured so that either loop can provide complete cooling coverage, eliminating the need for complex selective routing and simplifying the control logic despite the redundant hardware.
Solution Approach 2:
Switching valves act as intermediaries that manage the complexity of redundant cooling loops by providing simple on/off control between loops and storage nodes. These valves abstract the complexity of loop management, allowing automatic failover without requiring complex control systems or manual intervention.
3Use of energy by moving object
If cooling capacity is reduced during operation, then the energy consumption is lowered, but the reliability deteriorates due to insufficient cooling during failure conditions
Solution Approach 1:
The redundant cooling loop configuration ensures continuous full cooling capacity is always available to all storage nodes. When one loop is operational, it provides 100% cooling capacity; when both are operational, they share the load. This eliminates periods of insufficient cooling while maintaining energy efficiency through load sharing during normal operation.
Data Source
AI summary
A system for providing continuous cooling to a storage rack, including a storage rack, including a first storage node and a second storage node; a first cooling loop including a first cooling distribution unit (CDU); a second cooling loop including a second cooling distribution unit (CDU); a first inlet switching valve coupled between the first storage node and each of the first and the second CDUs; a second inlet switching valve coupled between the second storage node and each of the first and the second CDUs; wherein, when the first cooling loop experiences a failure, a state of the first inlet switch and a state of the second inlet switch is adjusted to i) prevent the first CDU from providing cooling of the first storage node and the second storage node, ii) allow only the second CDU to provide cooling of the first storage node and the second storage node.