RDMA Data Center Switching With Separate Protocol and Data Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network congestion methods fail to effectively consolidate protocol information across multiple packets, leading to inefficient resource usage and increased latency due to the need for separate processing of data and header portions, which cannot avoid congestion when destination paths are overwhelmed.
Innovation Solution
The method involves creating and transmitting separate protocol information (STPI) and data frames (DFoNP) separately, allowing switches to optimize resource usage and delay data frame processing until congestion is alleviated, while ensuring all necessary protocol information is maintained for efficient packet forwarding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If protocol information and data are transmitted separately (STPI and DFoNP), then network congestion is reduced and resource usage is optimized, but device complexity increases due to separate processing of data and header portions
Solution Approach 1:
The patent segments network packets into two separate components: STPI (Separately Transmitted Protocol Information) containing header and control data, and DFoNP (Data Frame or Network Packet) containing payload data. This segmentation allows switches to process and forward STPI independently while buffering DFoNP, thereby reducing congestion and optimizing resource utilization despite increased processing complexity.
Solution Approach 2:
The patent extracts protocol information from the traditional packet structure and transmits it separately via STPI. This extraction enables switches to handle control plane traffic independently from data plane traffic, improving network efficiency and resource management while requiring more complex switching logic to manage the separated components.
2Productivity
If separate protocol information transmission is implemented, then packet forwarding efficiency is improved, but latency increases due to separate processing of data and header portions
Solution Approach 1:
The patent implements preliminary action by transmitting STPI (protocol information) ahead of DFoNP (data frames). Switches can process routing decisions, quality of service policies, and other control functions based on STPI before the actual data arrives, enabling proactive resource allocation and congestion management that improves forwarding efficiency while minimizing latency through advance preparation.
Data Source
AI summary
A network system in a data center comprises of a plurality of interconnected nodes with network switches that enable the interconnection between the nodes that also include end nodes that comprise at least a processor and at least a memory. The end nodes are enabled to act as source nodes for data or destination nodes for data being transferred over the network system. Remote Direct memory access (RDMA) technology enable data to be accessed directly from a first memory address at the source node and transmitted over the interconnected network nodes to deliver the data to a memory address at the destination node without engaging the processors at either end nodes. Bypassing the processors reduce the latency associated with data retrieval, data transfer and data storage thereby improving the efficiency of the data center. The plurality of switches are configured for per-flow congestion control (PFC) to limit data loss.