Conversation-Sensitive Link Aggregation Frame Ordering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In link aggregation networks, maintaining frame ordering across conversations is challenging, especially in Distributed Resilient Network Interconnect (DRNI) configurations where multiple systems are involved, leading to potential frame duplication or reordering due to dynamic link redistribution.
Innovation Solution
Implementing a conversation-sensitive collection method where frames received on a link aggregation group port are associated with service and conversation identifiers, ensuring ordered sequences by discarding frames not allocated to the port, using a network device with a frame collector that compares conversation identifiers with port allocations and forwards frames accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If link aggregation is used to aggregate multiple links between nodes, then throughput and resiliency are improved, but frame ordering and sequence integrity deteriorate due to dynamic link redistribution
Solution Approach 1:
The patent segments the link aggregation group into multiple ports, each assigned to specific conversations. By dividing the LAG into port-conversation mappings, frames from each conversation are consistently routed through the same port, preventing reordering while maintaining aggregate throughput across multiple links.
Solution Approach 2:
The patent performs preliminary assignment of conversations to specific ports before frame transmission. The system pre-establishes which ports handle which conversations, so that when frames arrive, their destination port is already determined, ensuring ordered delivery without dynamic decision-making that could cause reordering.
2Reliability
If multiple nodes participate in LAG as virtual nodes, then system resiliency is improved, but frame duplication and reordering occur due to dynamic link redistribution
Solution Approach 1:
The patent applies local quality by assigning different conversation types to different ports based on their specific requirements. Each port develops specialized handling characteristics for its assigned conversations, ensuring that local frame ordering is maintained while the overall system benefits from multiple redundant paths.
Solution Approach 2:
The patent introduces port conversation allocation as an intermediary layer between the physical links and the conversations. This intermediary structure mediates frame routing decisions, ensuring that frames are directed to the correct port based on pre-established allocation rules, thereby preventing duplication and reordering in multi-node DRNI configurations.
3Productivity
If frames are dynamically redistributed across links, then load balancing is improved, but frame reordering and duplication occur
Solution Approach 1:
The patent introduces dynamic port conversation allocation that can adapt to changing network conditions. The system can dynamically adjust which ports handle which conversations based on link availability, performance metrics, or failure conditions, enabling load balancing while maintaining frame order through consistent port assignment for each conversation.
Solution Approach 2:
The patent implements feedback mechanisms that monitor frame delivery status and port performance. Based on this feedback, the system can adjust port conversation allocations to optimize load balancing while preventing frame reordering, ensuring that high-utilization paths are favored without compromising sequence integrity.
Data Source
AI summary
A method is executed by a network device for implementing conversation-sensitive collection for frames received on a port of a link of a link aggregation group. The network device executes an aggregator to collect the frames for aggregator clients, where each frame is associated with a service identifier and a conversation identifier. The service identifier identifies a data flow at a link level for a service. The conversation identifier identifies the data flow at a link aggregation group level, where each conversation data flow consists of an ordered sequence of frames, and where the conversation-sensitive collection maintains the ordered sequence by discarding frames of conversations not allocated to the port.