Hard Disk Backplane Isolation for Multi-Server Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-server systems face reliability issues due to failures in the hard disk backplane, where a single server node's failure can lead to storage failure across the entire system, affecting space utilization and data processing efficiency.
Innovation Solution
The architecture includes a hard disk backplane with isolated hard disk modules, each equipped with power and signal isolation units, and a power connection board that isolates power supplies and signal lines between modules, ensuring that failures in one server node do not affect others, and incorporates a CPLD module for monitoring and managing hard disk states, as well as a hot-plug assembly for protecting against signal noise during hot plug operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple servers share the same hard disk backplane to improve space utilization, then space utilization improves, but system reliability deteriorates because a single line failure causes whole system failure
Solution Approach 1:
The hard disk backplane is divided into multiple isolated hard disk modules, each with independent power supply and signal isolation units. This segmentation ensures that a failure in one module does not propagate to other modules, resolving the reliability issue while maintaining high space utilization through shared backplane infrastructure.
Solution Approach 2:
Signal isolation units and power supply isolation units are introduced as intermediary components between server nodes and hard disk modules. These intermediaries prevent direct failure propagation by electrically isolating different modules, thus improving reliability without sacrificing the space-efficient multi-server architecture.
2Device complexity
If server nodes are connected to hard disks through shared backplane lines, then device complexity is reduced, but reliability worsens due to single point of failure
Solution Approach 1:
The shared backplane connection is segmented into multiple isolated modules, each with its own power supply and signal isolation units. This maintains relatively simple connections within each module while preventing failure propagation, thus balancing connection simplicity with reliability.
Solution Approach 2:
Each hard disk module is designed with local isolation characteristics through dedicated power supply isolation units and signal isolation units. This local quality enhancement ensures that each module operates independently with improved reliability, while the overall system maintains manageable complexity through modular design.
3Device complexity
If power supplies are shared across hard disk modules, then device complexity is reduced, but reliability deteriorates because power failure in one module affects others
Solution Approach 1:
The power supply system is segmented into multiple independent power supply isolation units, each dedicated to a specific hard disk module. This segmentation increases reliability by preventing power failure propagation, while the modular design keeps overall power supply complexity manageable through standardized isolation circuits.
Solution Approach 2:
Power supply isolation units serve as intermediary components between the power connection board and individual hard disk modules. These intermediaries provide electrical isolation that prevents power failures in one module from affecting others, thus improving reliability without significantly increasing power supply complexity.
Data Source
AI summary
An architecture for improving reliability of a multi-server system is provided. The hard disk backplane is provided with at least two hard disk modules, each of which includes a power supply isolation unit and a signal isolation unit. The power connection board is connected to the power supply isolation unit in each of the at least two hard disk modules, so that power supplies of the at least two hard disk module are isolated from each other. Each of the at least two server nodes is connected to at least one of the hard disks through a corresponding signal isolation unit. The server node, the signal isolation unit and the hard disk which are connected to each other form an isolated data communication group. Signal isolation units in any two of isolated data communication groups belong to different hard disk modules.