Two-Stage Hash Distribution for Link Aggregation Bandwidth Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In link aggregation networks, physical lines often become underutilized due to identical hash calculation target information in frames, leading to concentration of frames on specific physical line ports, which results in inefficient use of communication bandwidth.
Innovation Solution
A communication apparatus that identifies frame flows, obtains hash rules based on flow identifiers, calculates hash values, and decides output ports for frames, allowing for the distribution of frames across multiple physical lines based on unique combinations of flow identifiers and hash values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If frames are divided among all physical lines using conventional hash calculation, then communication bandwidth utilization should improve, but frames with identical hash calculation target information still concentrate on specific physical line ports, resulting in underutilization of other lines
Solution Approach 1:
The patent segments the frame distribution process into two independent stages: first calculating a preliminary port group using conventional hash calculation on destination information, then further segmenting within that port group using a second hash calculation on flow identifier combinations. This two-stage segmentation ensures frames are distributed across all physical lines while maintaining flow consistency, resolving the contradiction between overall bandwidth utilization and effective frame distribution.
Solution Approach 2:
The patent introduces a new dimension to the hash calculation by combining flow identifier (FI) with destination information instead of using destination information alone. This dimensional expansion creates a more granular distribution key that effectively separates frames with identical original hash values, enabling better utilization of all physical line ports while maintaining communication efficiency.
2Device complexity
If frames are concentrated on fewer physical lines, then device complexity is reduced, but the communication band constructed using LAG technology is not used in an effective manner
Solution Approach 1:
The patent divides the frame distribution mechanism into two sequential segmentation steps: first segmenting frames into port groups using conventional hash calculation, then segmenting within each port group using flow identifier-based hash calculation. This segmented approach distributes frames across all physical lines effectively without requiring complex centralized control, thus improving LAG band utilization while keeping device complexity manageable.
Solution Approach 2:
The patent implements a self-service frame distribution mechanism where each switching element independently performs two-stage hash calculation to determine output ports. This eliminates the need for complex centralized arbitration or coordination between line interface units, achieving effective LAG utilization through simple, autonomous decision-making at each network element.
3Ease of manufacture
If conventional hash calculation is used for frame distribution, then implementation simplicity is maintained, but frames with identical destinations or hash values concentrate on the same physical line ports
Solution Approach 1:
The patent segments the hash calculation process into two simple, sequential steps that can be implemented using standard hashing hardware: first hash on destination information to get port group, then hash on flow identifier combination to get specific port. This segmented approach maintains implementation simplicity while effectively distributing frames across all physical line ports, avoiding concentration issues.
Solution Approach 2:
The patent adds a dimensional layer to the hash calculation by incorporating flow identifier into the second hashing stage. This dimensional enhancement maintains the simplicity of hash-based distribution while fundamentally improving port utilization by ensuring frames with identical destinations but different flows are distributed to different ports, achieving both simplicity and effectiveness.
Data Source
AI summary
A communication apparatus is connected to a network and identifies a flow of a frame input thereto and outputs the frame to a port. The communication apparatus includes: a flow identifying unit that obtains a flow identifier of the flow to which the frame belongs, and that obtains, based on the flow identifier, a hash rule corresponding to the flow identifier; a transfer destination controller that obtains by calculation a hash value of the frame by making use of the hash rule obtained by the flow identifying unit and that decides on an output port for the frame based on combination of the flow identifier and the hash value; and a frame output unit that outputs the frame to the output port decided by the transfer destination controller.