LBT over LAG Network Architecture for Throughput and Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network configurations in cloud computing environments face inefficiencies due to the mutual exclusivity of Load-Based Teaming (LBT) policies and Link Aggregation Group (LAG) methods, leading to suboptimal network throughput and availability, as they require selecting between LBT and LAG, making it impractical to switch between the two without significant reconfiguration overhead.
Innovation Solution
The implementation of an LBT over LAG network architecture that allows for the simultaneous use of both LBT policies and LAG methods, dynamically adjusting network topologies to optimize network throughput by binding a single dvport to multiple physical network interface cards (pNICs) and load balancing across multiple switches, thereby leveraging the attributes of both methods to enhance network performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If LBT policy is selected, then network throughput is improved, but network availability deteriorates due to mutual exclusivity with LAG methods
Solution Approach 1:
The patent combines LBT policy and LAG method into a unified network architecture where both can operate simultaneously. The system merges the load balancing capabilities of LBT with the redundancy benefits of LAG by allowing multiple pNICs to be bound to a single dvport while maintaining active-standby relationships, thus achieving both improved throughput and availability together rather than requiring mutual exclusion between the two methods
2Reliability
If LAG method is selected, then network availability is improved, but network throughput deteriorates due to mutual exclusivity with LBT policies
Solution Approach 1:
The patent merges LAG method and LBT policy into a single operational framework where the system can simultaneously maintain the active-standby redundancy of LAG for availability while applying load-based teaming for throughput optimization. The unified architecture allows pNICs to be organized in LAG groups for failover protection while also participating in LBT-based load distribution across multiple switches
3Productivity
If multiple pNICs are bound to a single dvport, then network throughput is improved, but device complexity increases
Solution Approach 1:
The patent introduces a network configuration manager as an intermediary component that automatically handles the complexity of binding multiple pNICs to a single dvport. This manager abstracts the complex configuration details from the operational system, providing automated policies for pNIC selection, LAG group management, and LBT parameter configuration, thus enabling throughput improvement without requiring manual management of the increased complexity
Data Source
AI summary
An example apparatus to manage network resources includes a link aggregator to: aggregate a first plurality of physical network interface cards to create a first link aggregated group, the first link aggregated group corresponding to a first virtual network interface card; and aggregate a second plurality of physical network interface cards to create a second link aggregated group, the second link aggregated group corresponding to a second virtual network interface card; and a link manager to: connect the first link aggregated group between a first distributed virtual port and a first top-of-rack switch; connect the second link aggregated group between a second distributed virtual port and a second top-of-rack switch; and remove an invalid connection between the first plurality of physical network interface cards and the second top-of-rack switch.