RAID Controller Logical Volume Migration via PCIe Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In Single Server High Availability (SSHA) systems, failover processes for RAID controllers are inefficient due to the need to read Disk Data Format (DDF) information from multiple storage devices, leading to delays in servicing host I/O requests during controller failures.
Innovation Solution
The implementation of shared Peripheral Component Interconnect Express (PCIe) memory access allows RAID controllers to quickly synchronize and migrate logical volumes by accessing DDF information from a PCIe Inbound Map, eliminating the need to read from individual storage devices and enabling faster failover.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If RAID controllers read DDF information from multiple storage devices during failover, then data availability is maintained, but failover time increases and host I/O requests are delayed
Solution Approach 1:
The patent implements preliminary action by having non-failed RAID controllers continuously synchronize and cache DDF information from storage devices in advance. This preliminary data gathering and caching ensures that when a controller failure occurs, the surviving controller already possesses the necessary DDF information to immediately assume control without needing to read from storage devices during the failover process, thus resolving the contradiction between maintaining data availability and reducing failover time.
Solution Approach 2:
The patent introduces an intermediary mechanism by implementing a shared memory space or communication channel between RAID controllers that facilitates rapid DDF information exchange. This intermediary allows controllers to share critical data without directly accessing storage devices during failover, enabling faster information transfer and reducing the time required for controller failure recovery while maintaining data availability.
2Device complexity
If RAID controllers use traditional I/O paths for failover, then system simplicity is maintained, but service disruption increases
Solution Approach 1:
The system performs preliminary actions by pre-synchronizing DDF information and maintaining cached copies of critical data structures in controller memory before failures occur. This preparation allows the surviving controller to immediately assume functionality without needing to perform complex real-time data retrieval operations during failover, thereby reducing service disruption time while keeping the overall system architecture relatively simple.
Solution Approach 2:
The patent implements dynamics by enabling RAID controllers to dynamically switch between different operational modes and data access paths based on system conditions. The controllers can dynamically access cached DDF information from shared memory or internal buffers rather than being forced to use traditional sequential storage access paths during failover, allowing flexible and optimized failure recovery without fundamentally changing system simplicity.
Data Source
AI summary
Methods and structure for migrating logical volumes are provided. The system includes a Redundant Array of Independent Disks controller, which includes a Peripheral Component Interconnect Express interface, a Serial Attached Small Computer System Interface port operable to communicate with another Redundant Array of Independent Disks controller, and a command unit. The command unit is able to direct the interface to access another Peripheral Component Interconnect Express interface at the other controller, to synchronize with Disk Data Format information from a Peripheral Component Interconnect Express Inbound Map of the other interface, to detect that the other controller has failed, and to utilize the Disk Data Format information to migrate a logical volume from the other controller to the controller.