Geographic Scaling in Container Cloud Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing container-based cloud infrastructures, such as Kubernetes, lack effective multi-tenancy support and struggle to dynamically scale resources across geographically distributed edge sites, particularly in meeting latency-sensitive service requirements.
Innovation Solution
A method and corresponding scaling entity for dynamically and automatically scaling container-based cloud infrastructure across multiple geographically distributed edge sites, by monitoring performance parameters and triggering amendments to infrastructure on other edge sites to meet performance and latency requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple geographically distributed edge sites are deployed to provide wide area coverage, then service coverage area is improved, but infrastructure cost and complexity increase
Solution Approach 1:
The system segments the edge cloud infrastructure into multiple geographically distributed sites, each running independent Kubernetes clusters. This allows service coverage to be expanded across multiple locations while maintaining independent, manageable infrastructure at each site, reducing overall system complexity through modular decomposition.
Solution Approach 2:
The patent introduces a new dimension of geographic distribution by deploying edge cloud sites across multiple locations rather than concentrating resources in a single data center. This spatial dimension enables wide area coverage while the federation mechanism manages complexity by providing unified control across distributed sites.
2Productivity
If cluster capacity is increased to satisfy performance requirements, then service performance is improved, but latency to remote users worsens
Solution Approach 1:
The system implements local quality by deploying service instances at multiple geographically distributed edge sites closer to end users. This ensures that users access services from the nearest edge site, minimizing latency while maintaining high performance through localized service delivery rather than relying on a single centralized cluster.
Solution Approach 2:
The patent resolves the latency-performance contradiction by adding a geographic dimension to service deployment. Instead of increasing capacity in one location, services are distributed across multiple geographic locations, allowing users to access services from nearby edge sites and thus reducing latency while maintaining performance.
3Productivity
If automated scaling is implemented for single cluster, then resource utilization is improved, but applicability to multi-cluster scenarios deteriorates
Solution Approach 1:
The patent creates a universal scaling mechanism that functions across both single-cluster and multi-cluster scenarios. The federation architecture enables the same scaling principles to be applied whether managing one cluster or multiple distributed clusters, making the system adaptable to different deployment scales and configurations.
Solution Approach 2:
The system merges the scaling capabilities of individual clusters into a unified federated scaling mechanism. This combines the resource utilization benefits of automated scaling with the geographic distribution advantages of multi-cluster deployment, enabling coordinated scaling across the entire federated edge cloud infrastructure.
4Reliability
If separate Kubernetes clusters are deployed for each tenant to provide multi-tenancy, then tenant isolation is improved, but infrastructure cost and complexity increase
Solution Approach 1:
The system segments the edge cloud infrastructure into multiple geographically distributed sites with independent Kubernetes clusters, providing tenant isolation through spatial separation. This segmentation approach maintains strong multi-tenancy boundaries while reducing overall complexity by distributing infrastructure across independent locations rather than consolidating in a single complex system.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The application relates to a method for operating a first scaling entity 100 of a container based cloud infrastructure distributed over a plurality of geographically distributed edge sites 20-22 of a network, wherein the container based cloud infrastructure provides at least one service to a user of the network, the method comprising: the step of determining at least one performance parameter influencing a performance how a first part of the container based cloud infrastructure provided on a first edge site of the plurality of edge sites where the first scaling entity is located provides the service to the user. Based on the determined at least one performance parameter, it is determined whether a scaling of the cloud infrastructure located on at least one other edge site of the plurality of geographically distributed edge sites outside the first edge site is necessary, wherein in the affirmative, an amendment of the container based cloud infrastructure is determined at the at least one other edge site which is configured to provide the service to the user. The determined amendment of the container cloud infrastructure is triggered at the at least one other edge site.