Fabric Adapter Packet Segmentation for Network Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems in data centers face bottlenecks due to limited network and system I/O bandwidth, as well as processing constraints of protocol algorithms, which hinder efficient data transmission and processing across multiple processors and networks.
Innovation Solution
The proposed system employs a Server Fabric Adapter (SFA) communication system that separates network packets into headers and payloads, forwarding headers to controlling hosts and payloads in parallel through a bundled interface to a global memory pool using scatter gather lists (SGLs).
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If network packets are processed using traditional I/O bandwidth and protocol algorithms, then data transmission can be achieved, but system bottlenecks occur and communication speed is reduced
Solution Approach 1:
The patent segments network packets into separate header and payload portions, allowing them to be processed through different pathways simultaneously. Headers are forwarded to controlling hosts for protocol processing while payloads are directly routed to memory buffers via scatter gather lists, eliminating the bottleneck of sequential processing and improving overall communication speed and system throughput.
2Speed
If protocol algorithms are processed in software, then network packets can be handled, but processing speed is slowed down
Solution Approach 1:
The patent extracts the payload processing from the software-based protocol algorithm path and creates a separate direct memory access pathway using scatter gather lists. This allows payload data to bypass the software processing bottleneck and be directly transferred to memory buffers, significantly improving packet processing speed while reducing the computational complexity burden on the CPU.
3Quantity of substance
If I/O bandwidth is increased, then data transmission capacity is improved, but system bottlenecks persist due to processing limitations
Solution Approach 1:
The patent introduces a new dimensional pathway for data transmission by implementing parallel processing channels: one channel handles headers through the traditional I/O bandwidth path to controlling hosts, while another channel handles payloads through direct memory access via scatter gather lists. This multi-dimensional approach allows the system to simultaneously increase data transmission volume and maintain high transmission speed by avoiding sequential processing bottlenecks.
Data Source
AI summary
In an aspect, a system for an optimally balanced networked system is disclosed. The system includes a fabric adapter communication system communicatively coupled to a plurality of network ports and a plurality of controlling hosts. The fabric adapter communication system is configured to receive a network packet from, or transmit a network packet to, a network port of the plurality of network ports. The fabric adapter communication system is configured to separate the network packet into different portions, each portion including a header or a payload. The fabric adapter communication system is configured to forward the headers of the different portions to one or more controlling hosts. The fabric adapter communication system is configured to forward multiple payloads of the different portions in parallel through a bundled interface to multiple memory buffers of a global memory pool based on one or more scatter gather lists (SGLs).