Network Controller for Container Workload Visibility in Blade Switches
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data center environments with blade switches, existing technologies face challenges in maintaining network visibility for containerized workloads, particularly in complex integrated configurations, where the additional levels of indirection obscure the visibility of traffic and connections, making it difficult to monitor and debug issues effectively.
Innovation Solution
A network controller determines the physical network topology, collects virtual machine and container workload information, and generates an integrated correlated visualization mapping, providing insights into containerized workloads without requiring custom agents on servers or switches, by integrating data from virtual machine managers and container orchestrators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If containerized deployments are implemented in data center environments with blade switches, then workload flexibility and resource utilization are improved, but network visibility and monitoring capabilities deteriorate due to additional levels of indirection
Solution Approach 1:
The patent introduces a network controller as an intermediary component that collects information from multiple sources (physical network topology, virtual machine manager, blade system management software, container orchestrator) and generates an integrated correlated visualization mapping. This mediator reconciles the indirection layers between containerized workloads and physical network infrastructure, restoring visibility without compromising workload flexibility.
Solution Approach 2:
The patent adds a new dimensional layer - the integrated correlated visualization mapping - that correlates container workload information with physical network topology across multiple indirection layers. This additional dimension enables visibility through the complexity by mapping virtual container entities to their physical network representations, allowing monitoring without changing the containerized deployment model.
2Productivity
If integrated data center configurations with blade switches are used, then space utilization and infrastructure efficiency are improved, but complexity of monitoring and debugging increases
Solution Approach 1:
The patent merges multiple monitoring data sources (physical network topology from network controllers, virtual machine topology from VM managers, blade system information from management software, and container workload data from orchestrators) into a single integrated correlated visualization mapping. This consolidation reduces monitoring complexity by providing a unified view despite the integrated infrastructure's complexity.
Solution Approach 2:
The patent segments the monitoring function into a dedicated network controller that independently collects and processes information from various infrastructure components. This segmentation allows the monitoring system to handle complexity systematically by processing each data type separately before integrating them, making the monitoring process more manageable despite infrastructure integration.
3Adaptability or versatility
If multiple levels of indirection are introduced in containerized deployments, then abstraction and portability are improved, but difficulty in tracing traffic and connections increases
Solution Approach 1:
The network controller acts as an intermediary that bridges the abstraction layers. It collects container workload information from the container orchestrator and correlates it with physical network topology data, creating a mapping that traces traffic paths through multiple indirection layers while preserving the portability benefits of container abstraction.
Solution Approach 2:
The patent creates a virtual representation or copy of the physical network topology correlated with container workload information. This copied mapping enables traffic tracing and connection analysis without interfering with the actual containerized deployment architecture, maintaining portability while enabling measurement through the indirection layers.
Data Source
AI summary
In one embodiment, an illustrative method herein may comprise: determining, by a network controller, physical network topology of a data center network; collecting, by the network controller, virtual machine related network topology of the data center network from a virtual machine manager for the data center network; collecting, by the network controller, virtual ethernet flow mapping information of the data center network from a blade system management software for the data center network; collecting, by the network controller, container workload information of the data center network from a container orchestrator for the data center network; and generating, by the network controller, an integrated correlated visualization mapping of containers and associated network attachment points in the data center network based on the physical network topology, the virtual machine related network topology, the virtual ethernet flow mapping information, and the container workload information.