Cloud Load Balancing via Dynamic Pricing Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current load-balancing systems for cloud services face challenges in economically optimizing the distribution of web services across non-homogeneous cloud providers, as they struggle to efficiently manage dynamic pricing models and traffic patterns.
Innovation Solution
A load-balancing system that utilizes a processor-driven algorithm to compare static and real-time pricing data from multiple cloud vendors, combined with traffic profiling, to determine the most economical cloud provider or combination of providers for balancing traffic, leveraging REST APIs for data collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If traditional load-balancing systems are used, then traffic distribution is maintained, but cost optimization across multiple cloud providers is insufficient
Solution Approach 1:
The patent introduces a cloud broker as an intermediary component that sits between the load balancer and multiple cloud providers. This broker collects pricing information from various providers, processes it according to pricing policies, and guides traffic routing decisions. By placing this intermediary layer, the system achieves cost optimization without requiring complex integration logic in each load-balancing component, thus resolving the contradiction between cost efficiency and system complexity.
Solution Approach 2:
The system performs preliminary actions by collecting and storing pricing information from multiple cloud providers in advance, before actual traffic routing decisions are needed. The cloud broker maintains updated pricing data and pre-processes cost calculations, so that when traffic needs to be routed, the decision can be made quickly based on pre-analyzed cost data. This preliminary preparation enables cost optimization without adding real-time computational complexity to the routing decision process.
2Loss of energy
If static pricing data is used for load balancing, then system simplicity is maintained, but real-time cost optimization is achieved
Solution Approach 1:
The cloud broker continuously collects pricing information from multiple cloud providers in the background, maintaining an always-updated view of current pricing conditions. This continuous data collection ensures that when traffic routing decisions are made, the most current pricing information is available without requiring time-consuming data gathering at the moment of decision. The continuous action of data collection resolves the contradiction between achieving real-time cost optimization and the time required to gather necessary pricing data.
3Reliability
If traffic is distributed across non-homogeneous cloud providers, then vendor independence and redundancy are improved, but economic optimization becomes more difficult
Solution Approach 1:
The cloud broker is designed with universal functionality that can handle pricing information from multiple different cloud providers with varying pricing models, structures, and APIs. It provides a unified interface for collecting, normalizing, and processing pricing data from heterogeneous sources, and applies consistent pricing policies across all providers. This multi-functional capability enables the system to maintain vendor independence and redundancy while simplifying pricing management through a single universal component.
Data Source
AI summary
The disclosed technology includes accessing a first network application programming interface exposed by a first cloud provider of the plurality of cloud providers to identify a first pricing profile, the first pricing profile associated with the first Cloud provider. Upon identifying the first pricing profile, accessing a second network application programming interface exposed by a second cloud provider of the plurality of cloud providers to identify a second pricing profile, the second pricing profile associated with the second Cloud provider. A load balancing decision is determined comparing the identified first pricing profile with the identified second pricing profile. Next, the determined load balancing decision is executed on a monitored computing-traffic.