Workload Placement via Switch Controller Network Throughput Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current workload deployment methods in software-defined networks (SDNs) often lead to network congestion due to insufficient visibility into optimal placement, as they primarily consider localized server utilization and I/O requirements without accounting for global network congestion.
Innovation Solution
A system and method where a switch controller communicates with network devices to determine the throughput associated with a new or current workload and selects an optimum location for placement based on this throughput, using statistics from network devices to predict and avoid congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If workload deployment considers only localized server utilization and I/O requirements, then deployment simplicity is maintained, but network congestion occurs due to insufficient visibility of global network conditions
Solution Approach 1:
The patent transitions workload deployment from a localized two-dimensional view (server utilization and I/O requirements) to a global multi-dimensional view that incorporates network throughput, congestion conditions, and flow statistics across the entire data center network. This dimensional expansion enables the system to identify optimal deployment locations that avoid network congestion while maintaining deployment automation.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring network flow statistics, throughput metrics, and congestion conditions across the data center network. This real-time feedback information is fed back into the workload deployment decision-making process, allowing the system to dynamically adjust deployment choices based on current network conditions, thereby preventing congestion while maintaining operational simplicity.
2Productivity
If global network throughput is considered for workload placement, then network performance is improved, but system complexity increases due to enhanced monitoring and analysis requirements
Solution Approach 1:
The patent employs a universal workload placement framework that integrates multiple functions into a single system: network traffic monitoring, throughput calculation, congestion detection, and optimal placement decision-making. This multi-functional approach consolidates what would otherwise require separate systems, maintaining manageable complexity while achieving global network performance optimization through unified control.
Solution Approach 2:
The system introduces an intermediary workload placement optimization layer that sits between the workload management system and the physical network infrastructure. This intermediary component abstracts the complexity of global network monitoring and analysis, providing simplified deployment decisions while leveraging detailed network statistics, thereby improving network performance without proportionally increasing overall system complexity.
3Measurement precision
If real-time network statistics are collected from all network devices, then accurate throughput determination is achieved, but information processing overhead increases
Solution Approach 1:
The patent applies partial action by selectively collecting and processing network statistics only from devices and paths relevant to the specific workload deployment decision. Rather than continuously processing all network data, the system gathers throughput and flow statistics针对性地 (targetedly) for the paths affected by the workload placement under consideration, achieving accurate throughput determination while minimizing unnecessary information processing overhead.
Data Source
AI summary
In one embodiment, a system includes a switch controller configured to communicate with each of a plurality of network devices in a network, the switch controller including a processor and logic integrated with and/or executable by the processor, the logic being configured to: receive a request to instantiate a new workload or to move a current workload in the network, determine a throughput associated with the new or current workload, and select an optimum location to place the new workload or to move the current workload in the network based on the throughput associated with the new or current workload. According to another embodiment, a method includes receiving a request at a switch controller to instantiate or move a workload, determining a throughput associated with the workload, and selecting an optimum location to place or move the workload based on the throughput associated with the workload.