Integrated Gateway Service for Multi-Infrastructure SDN Reconciliation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software-defined networking (SDN) architecture systems face challenges in cloud-native adoption due to complexity in lifecycle management, scale limitations in configuration modules, and the lack of a command-line interface (CLI)-based interface, making it difficult to seamlessly integrate and manage virtualized compute infrastructures such as virtual machines and containers across diverse environments.
Innovation Solution
A software-defined networking (SDN) architecture system that implements an integrated gateway service (GSI) to model and abstract underlying compute infrastructures, using a custom controller to understand infrastructure-specific APIs and manage network resources, including a containerized network router (CNR) for high-performance networking, and automate configuration and lifecycle management across various compute environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If an integrated gateway service is implemented to abstract compute infrastructures, then ease of operation is improved, but device complexity increases
Solution Approach 1:
The gateway service instance acts as an intermediary layer between the orchestration platform and diverse compute infrastructures (virtual machines, containers, bare metal). It provides a unified abstraction interface that translates between different infrastructure-specific APIs and the orchestration platform, simplifying operations while managing complexity internally through standardized resource models and infrastructure-agnostic workflows.
Solution Approach 2:
The gateway service instance implements multi-functional capabilities to handle diverse compute infrastructures through a single unified interface. It can manage virtual machines, containers, and bare metal servers simultaneously, providing universal networking, storage, and compute resource orchestration across heterogeneous environments without requiring infrastructure-specific configuration workflows.
2Adaptability or versatility
If custom controllers are used to manage infrastructure-specific APIs, then adaptability is improved, but device complexity increases
Solution Approach 1:
The control functionality is segmented into independent custom controllers, each responsible for managing specific infrastructure types (e.g., one controller for virtual machines, another for containers). Each controller implements infrastructure-specific API logic independently, allowing the system to adapt to diverse compute infrastructures while keeping individual controller complexities manageable and modular.
Solution Approach 2:
Custom controllers serve as intermediary components between the orchestration platform and infrastructure-specific systems. They translate orchestration platform requests into infrastructure-specific API calls and vice versa, enabling adaptability to different compute infrastructures while shielding the core orchestration platform from infrastructure complexity.
3Reliability
If gateway service instance reconciles infrastructure state, then reliability is improved, but loss of time increases
Solution Approach 1:
The gateway service instance continuously monitors and reconciles the actual state of compute infrastructure with the desired state defined by the orchestration platform. It implements feedback loops that detect state deviations (e.g., network connectivity issues, resource allocation mismatches) and automatically correct them, ensuring reliable network connectivity while minimizing manual intervention time through automated state verification and repair.
Data Source
AI summary
In general, techniques are described for a SDN architecture system that implements an integrated gateway service. In an example, network controller comprises processing circuitry and memory and is configured to: configure a virtual network in a cluster of nodes of a compute infrastructure, the cluster of nodes managed in part by the network controller; receive a manifest for a gateway service instance that abstracts a transit gateway resource configurable in a plurality of different types of compute infrastructures, wherein the manifest specifies an intended state of a transit gateway object of the gateway service instance; and reconcile the intended state of the transit gateway object for the gateway service instance by sending configuration data, generated based on the transit gateway object, to an interface for the compute infrastructure to configure a transit gateway to forward network packets between a compute infrastructure node of the compute infrastructure and the virtual network.