Fibre Channel Host Bus Adapter Reset for High Availability Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current high availability data storage systems using Fibre Channel technology face challenges in quickly failing over FC connections during operating system panic conditions, as existing methods are either too slow, disrupt disk dump processes, or require complex firmware changes.
Innovation Solution
Implementing a method to rapidly reset FC host bus adapter devices during OS panic, allowing for immediate failover of FC connections while maintaining disk dump operations, using PCIe function level reset (FLR) and managing reset policies through HA management modules, which is scalable and firmware-independent.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system waits for OS panic to finish and then reboots or shuts down, then the system can recover from panic condition, but the failover time becomes very long (minutes or hours)
Solution Approach 1:
The patent applies preliminary action by detecting the panic condition and initiating the FC device reset process before the OS disk dump completes. The system identifies panicking processes and resets FC HBA devices proactively, allowing failover to occur in milliseconds rather than waiting for the lengthy disk dump process to finish naturally.
2Loss of time
If the system forces a system reset to interrupt OS disk dump process, then the failover time is reduced, but the disk dump information is lost
Solution Approach 1:
The patent segments the system recovery process into two independent parts: (1) resetting FC HBA devices to enable fast failover, and (2) allowing the OS disk dump process to continue independently. This segmentation allows the FC reset to occur without interrupting the disk dump, preserving crash dump information while achieving rapid failover.
3Loss of time
If firmware is used to reset the device, then the reset can occur quickly, but the firmware complexity increases and maintenance becomes harder
Solution Approach 1:
The patent substitutes firmware-based reset mechanisms with an OS-level solution using kernel modules and device drivers. The panic detection and FC device reset functionality is implemented in the OS kernel space, replacing the need for complex firmware modifications. This approach maintains fast reset capability while simplifying maintenance and updates.
4Adaptability or versatility
If device specific reset functionality is implemented, then the reset can be targeted to specific devices, but changes are required for each driver and hardware support is needed
Solution Approach 1:
The patent implements a universal FC device reset mechanism that works across different FC HBA devices and operating systems through standard PCIe function level reset (FLR) operations. The kernel module detects panic conditions and applies generic reset procedures to FC devices without requiring device-specific driver modifications or hardware changes, achieving adaptability through universality.
Data Source
AI summary
Embodiments for systems and methods of resetting network devices for failover operation, by receiving an operating system panic function call, disabling error reporting for all device ports of a device to be reset to prevent primary interrupts to a disk dump operation, performing a function level reset of the all the device ports to be reset, and performing the reset operation on the device in a single thread context to prevent secondary interrupts to the disk dump process.