Network-Aware Global Load Balancing for Data Center Resource Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current load balancing decisions in network carriers are made without adequate consideration for underlying network resources, leading to suboptimal application and network resource utilization, as they lack awareness of network usage, capabilities, and congestion status, resulting in reduced application performance and failure to meet service objectives.
Innovation Solution
A network-aware global load balancing system that associates with local and remote data centers to select servers and routes based on application and network resource information, using cross-stratum optimization to integrate application and network stratum resources for informed decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If load balancing decisions are made without network resource awareness, then device complexity is reduced, but application performance deteriorates and network resource utilization becomes suboptimal
Solution Approach 1:
The patent introduces a network awareness component as an intermediary between the load balancer and network resources. This component collects network status information (congestion, capabilities, usage) and provides it to the load balancing decision-making process, enabling informed decisions without significantly increasing overall system complexity.
Solution Approach 2:
The system implements feedback mechanisms where network status information is continuously collected and fed back to the load balancing component. This feedback loop enables dynamic adjustment of load balancing decisions based on current network conditions, improving application performance while maintaining manageable complexity through automated control.
2Productivity
If network resource awareness is integrated into load balancing, then network resource utilization improves, but device complexity increases
Solution Approach 1:
The patent segments the load balancing system into distinct functional components: a network awareness component that collects and processes network information, and a decision-making component that uses this information. This segmentation allows each component to specialize in specific tasks, improving network resource utilization while managing overall system complexity through modular design.
3Reliability
If comprehensive network and application resource information is collected, then service objectives are met, but loss of information increases due to communication overhead
Solution Approach 1:
The patent implements partial information collection by focusing on specific network parameters most relevant to load balancing decisions (congestion status, path capabilities, current usage). Rather than collecting all possible network information, the system selectively gathers only the necessary data, reducing communication overhead and information loss while still meeting service objectives.
Data Source
AI summary
An apparatus comprising a network aware (NA) global load balancing (GLB) component configured to associate with a local data center (DC), couple to an end user via a network, select a server in the DC or in one or more remote DCs coupled to the network based on application resource information and network resource information, and select a route in the network based on the application resource information and the network resource information.