Switch Fabric Routing Using Global Load Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network architectures face challenges in scalability, versatility, and efficiency due to increasing network load and diverse traffic types from applications like high-performance computing and Internet of Things, with conventional congestion control mechanisms leading to slow responses and inefficient network utilization.
Innovation Solution
A network architecture with flow channels and adaptive routing, utilizing software-generated global load information and distributed congestion control, allows for dynamic flow setup and teardown, reducing latency and improving network performance by up to three orders of magnitude.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional congestion control mechanisms are used, then network stability is maintained, but network response speed and utilization efficiency deteriorate
Solution Approach 1:
The patent implements preliminary action by proactively establishing flow channels and pre-determining routing paths before congestion occurs. The system uses software-generated global load information to predict network conditions and set up flow channels in advance, allowing the network to respond faster to traffic demands without waiting for conventional congestion signals. This pre-positioning of resources resolves the contradiction by enabling quick response while maintaining stability through planned resource allocation.
Solution Approach 2:
The patent applies feedback mechanisms by continuously monitoring network load conditions and using software-generated global load information to dynamically adjust flow channel establishment and routing decisions. The system receives feedback about network state from multiple sources, processes this information centrally, and uses it to make real-time routing adjustments. This closed-loop feedback system enables fast response to changing conditions while maintaining network stability through informed decision-making.
2Adaptability or versatility
If network architecture supports diverse traffic types and applications, then versatility is improved, but device complexity and management difficulty increase
Solution Approach 1:
The patent implements universality by creating a flow channel mechanism that can handle multiple types of traffic (HPC, media streaming, IoT, etc.) through a single unified framework. The software-generated global load information serves all traffic types universally, and the flow channel establishment process works the same way regardless of traffic type. This multi-functional approach allows the network to support diverse applications without requiring separate specialized mechanisms for each traffic type, thus improving versatility while controlling complexity.
Solution Approach 2:
The patent introduces an intermediary layer (the software-generated global load information and centralized control mechanism) that mediates between diverse traffic sources and the network infrastructure. This intermediary processes routing decisions for all traffic types, abstracting the complexity of managing diverse applications away from individual network components. By centralizing the intelligence in this intermediary layer, the system can handle versatile traffic types without increasing the complexity of individual network devices.
3Productivity
If dynamic flow setup and teardown is implemented, then network efficiency and latency are improved, but control complexity and overhead increase
Solution Approach 1:
The patent applies self-service by enabling flow channels to be automatically established and torn down based on real-time network conditions without manual intervention. The software-generated global load information drives automatic flow channel creation when needed and removal when no longer necessary. This self-managing capability allows the network to dynamically adapt to changing traffic patterns, improving efficiency and reducing latency while avoiding the complexity of manual flow management. The system serves itself by making autonomous routing decisions based on monitored conditions.
Data Source
AI summary
Systems and methods are provided for performing routing in a switch network or fabric. Switches can be configured in a hierarchical topology having a plurality of groups, where switches in a group are connected to one another, and groups are connected to other groups. Routing can be performed by maintaining per-group group load information. A packet can be routed between at least two groups using the per-group group load information to effect a set of routing decisions. The set of routing decisions can be biased towards or away one or more paths.