Stateful Offload Adapter Priming Failover Mechanism
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern communication adapters with stateful offload capabilities face challenges in providing high availability and transparent failover due to the loss of offloaded state when experiencing an outage, as existing mechanisms like Automatic Path Migration (APM) are limited to finding alternate paths within the same adapter or host channel adapter and cannot handle simultaneous failures across all ports.
Innovation Solution
A method and system for performing a seamless failover from a primary adapter to a secondary adapter by detecting imminent failure, initiating a failover priming operation, and migrating ingress and egress traffic, allowing the secondary adapter to handle data processing without host system intervention, thereby maintaining state information and minimizing downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Automatic Path Migration (APM) is used to find alternate paths, then path redundancy is improved, but failover capability is limited to same adapter ports only
Solution Approach 1:
The system segments the failover capability by separating path migration (handled by APM within an adapter) from adapter-level failover (handled by the host system). This allows each component to specialize: APM handles port-level redundancy while the host system coordinates cross-adapter failover, resolving the limitation where APM could only provide same-adapter path redundancy.
Solution Approach 2:
The host system acts as an intermediary between multiple adapters, coordinating failover operations across adapter boundaries. The host system receives failure notifications from adapters, selects appropriate standby adapters, and manages the failover process, enabling versatility beyond what single-adapter APM mechanisms can achieve.
2Speed
If stateful offload is implemented in communication adapters, then processing latency is reduced, but state information is lost during adapter failure
Solution Approach 1:
The system implements feedback mechanisms where adapters notify the host system of their operational status and state information. The host system continuously monitors adapter health and maintains awareness of offloaded state, enabling it to make informed failover decisions and restore state information to standby adapters, thus preserving state during failures while maintaining low-latency processing.
Solution Approach 2:
The host system performs preliminary actions by pre-configuring standby adapters with necessary state information and failover capabilities before failures occur. This includes pre-establishing communication channels, preparing buffer configurations, and maintaining state snapshots, so that when a failure occurs, the transition to standby adapters can happen quickly without state loss.
3Loss of time
If failover priming operation is performed, then failover time is reduced, but adapter configuration complexity increases
Solution Approach 1:
The failover priming operation employs self-service mechanisms where the host system automatically performs configuration tasks on standby adapters without manual intervention. The host system programmatically configures buffer parameters, establishes communication channels, and prepares state information, reducing failover time while managing complexity through automation rather than manual configuration procedures.
Data Source
AI summary
Mechanisms, in a data processing system comprising a first adapter and second adapter, for performing a failover operation from the first adapter to the second adapter are provided. The mechanisms detect that an imminent failure of the first adapter is likely to occur and initiate a failover priming operation in the first adapter and second adapter in response to detecting the imminent failure. The failover priming operation configures ingress and egress buffers of the second adapter to have a similar configuration to ingress and egress buffers of the first adapter. The mechanisms migrate processing of ingress data traffic to the second adapter prior to failure of the first adapter such that the first adapter processes egress data traffic from the data processing system and the second adapter processes ingress data traffic to the data processing system.