Network Switch Port Aggregation via Trunk Header Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network switching approaches incur high computational overhead due to message classification processing at each switch, and the trunk header destination becomes moot in static forwarding configurations, leading to inefficiencies in handling increased bandwidth demands.
Innovation Solution
The solution involves disabling routing logic in network switches by storing packet classification information in the trunk header fields, allowing direct forwarding to an aggregation port, which connects to a packet processor, thereby avoiding redundant processing and enabling efficient interconnection of switches for enhanced throughput.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If message classification processing is performed at each switch in conventional stacked switch configurations, then routing control is maintained, but computational overhead increases and processing efficiency decreases
Solution Approach 1:
The patent extracts the message classification processing function from individual switches and consolidates it at the packet processor. The trunk header fields that would normally contain routing control information are repurposed to carry classification information to the packet processor, which then performs the classification processing centrally rather than at each switch, thereby reducing computational overhead at individual switches while maintaining overall routing control.
2Loss of information
If trunk header destination fields are used for routing control in static forwarding configurations, then routing accuracy is maintained, but the fields become redundant and waste bandwidth
Solution Approach 1:
The patent makes the trunk header fields multi-functional by using them to carry classification information rather than routing control information in static forwarding configurations. Since the output port is already known in static forwarding, the trunk header destination fields serve a different purpose - transporting classification information to the packet processor - thereby eliminating redundancy and improving bandwidth utilization.
3Productivity
If switches interconnect via aggregation ports with static forwarding, then bandwidth handling is improved, but routing flexibility is reduced
Solution Approach 1:
The patent introduces the packet processor as an intermediary that receives packets from aggregation ports with classification information in the trunk headers. The packet processor performs the routing decisions centrally based on the classification information, thereby maintaining routing flexibility despite the static forwarding configuration at individual switches. This intermediary approach allows switches to focus on high-speed packet forwarding while the packet processor handles routing logic.
Data Source
AI summary
A network switch configures a static forwarding to a packet processor by suppressing packet switching and forwards all traffic received on a group of ports a trunk port for aggregation. A trunk header is overloaded with message classification information for use at the downstream packet processor. Routing logic retrieves the packet classification information and stores the information in control fields that are ignored due to the static forwarding and local switching disablement. The static forwarding forwards the packet, with the appended classification information, to a packet processor via the aggregation port. Packet classification information is indicative of the type of the message traffic and is performed upon packet arrival at the switching device. The packet processor reads the classification information from the overloaded control fields, rather then expending processing resources to determine the classification, and sends the message packet on an ingress port to a switching fabric for further transport.