Virtual HBA Port Migration for Storage Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtualization services platforms, physical HBA failures lead to storage device inaccessibility for virtual machine guests, and existing multipathing solutions are inefficient, causing redundant virtual HBA port overheads and performance issues during storage area network changes.
Innovation Solution
Implementing a system with a primary and backup physical HBA, where virtual HBA ports are migrated from the primary to the backup HBA upon failure, with incoming I/O requests placed in a wait queue and subsequently transferred to an active queue for processing, eliminating the need for redundant WWNs and login sessions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multipathing is implemented to mitigate physical HBA failure, then storage device accessibility is improved, but physical HBA resource efficiency deteriorates due to redundant virtual HBA ports and login sessions
Solution Approach 1:
The patent extracts the essential function of redundancy (storage accessibility) from the inefficient multipathing implementation. Instead of maintaining multiple active virtual HBA ports and login sessions, the system uses a single active path with a standby path that is activated only when needed, removing the wasteful redundant resources while preserving the reliability function
Solution Approach 2:
The system dynamically transitions between single-path and multi-path configurations. The virtual HBA ports and login sessions are dynamically created or activated on the standby physical HBA only when a failure is detected, rather than maintaining them statically all the time. This dynamic approach optimizes resource efficiency while ensuring reliability
2Reliability
If multiple alternate virtual HBA ports are maintained for multipathing, then storage device accessibility is improved, but system complexity increases due to additional RSCN processing overhead
Solution Approach 1:
The patent removes the unnecessary complexity of processing RSCN notifications on multiple alternate virtual HBA ports. By maintaining only one active path at a time, the system extracts the essential redundancy function while eliminating the harmful RSCN processing overhead that occurs when multiple ports are actively registered with storage devices
3Reliability
If virtual HBA ports are migrated from primary to backup physical HBA, then reliability is improved, but migration time may increase storage access latency
Solution Approach 1:
The system performs preliminary actions by pre-configuring the standby physical HBA with the necessary virtual HBA ports and login sessions before a failure occurs. When a failure is detected, the migration is essentially instantaneous because the backup path is already prepared and ready to activate, minimizing migration time and storage access latency
Solution Approach 2:
The patent implements a rapid failover mechanism that rushes through the migration process by directly switching to the pre-prepared standby path. The system skips lengthy reconfiguration steps by having the backup virtual HBA ports and login sessions already established, enabling quick activation that minimizes disruption to storage access
Data Source
AI summary
Examples disclosed herein relate to virtual Fiber Channel port migration. In some examples disclosed herein, a failure in a primary physical host bus adapter (HBA) hosting a plurality of virtual HBA ports assigned to a virtual machine may be detected. In response to the detecting, incoming input/output (I/O) requests from the virtual machine may be placed into a wait queue, and the plurality of virtual HBA ports may be migrated to a backup physical HBA. After the plurality of virtual HBA ports have been migrated to the backup physical HBA, the I/O requests in the wait queue may be transferred to an active queue to be processed using the plurality of virtual HBA ports registered on the backup physical HBA.