SR-IOV Virtual Function Failover via Hypervisor Standby Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data centers providing cloud services, existing virtualization technologies face challenges in minimizing the impact of hardware failures or updates on SR-IOV adapters, leading to communication losses between logical partitions and virtual functions.
Innovation Solution
The implementation of a failover mechanism by a hypervisor, which instantiates a standby virtual function, detects communication loss, places both active and standby virtual functions in an error state, remaps the logical partition to the standby virtual function, and transitions it into an error recovery state, ensuring seamless data communications without modifying I/O parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If SR-IOV adapter is used for virtualization services, then I/O performance and network bandwidth are improved, but system reliability deteriorates due to hardware failures or updates causing communication loss
Solution Approach 1:
The patent instantiates a standby virtual function in advance before any failure occurs. This standby VF is prepared and ready to immediately take over if the active VF fails, eliminating the need for recovery actions after failure and preventing communication loss entirely.
Solution Approach 2:
The patent changes the state parameters of virtual functions by transitioning them between active, standby, error, and error recovery states. This state management allows the system to dynamically adjust VF roles based on operational conditions, maintaining reliability while preserving I/O performance.
2Reliability
If failover mechanism is implemented, then system reliability is improved, but device complexity increases due to additional standby virtual function management
Solution Approach 1:
The patent implements automatic failover detection and execution without requiring external intervention. The system monitors communication status and automatically transitions from active to standby VF when failure is detected, reducing operational complexity despite the added architectural elements.
Solution Approach 2:
The patent creates a standby virtual function that is essentially a copy or duplicate of the active VF. This copying approach simplifies the failover mechanism because the standby VF already has the necessary configuration and state, requiring only activation rather than complex reconstruction during failure.
3Reliability
If communication loss is detected and failover is executed, then system reliability is improved, but loss of time occurs during the failover transition process
Solution Approach 1:
The standby virtual function is instantiated and configured in advance, with all necessary resources and mappings prepared beforehand. When failover is needed, the transition is immediate rather than requiring setup time, thus minimizing or eliminating loss of time during the failover process.
Solution Approach 2:
The patent implements a direct failover mechanism that skips intermediate recovery steps by having the standby VF ready to immediately assume the active role. This rushing through the failover process minimizes the time window where communication could be lost.
Data Source
AI summary
Failover of a virtual function exposed by an SR-IOV adapter of a computing system, including: instantiating, by a hypervisor, a standby virtual function in the computing system; detecting a loss of communication between a logical partition and an active virtual function mapped to the logical partition; placing the active virtual function and the standby virtual function in an error state; remapping the logical partition to the standby virtual function; and placing the standby virtual function in an error recovery state.