Edge Node Application Deployment via On-Demand Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud computing systems lack efficient on-demand application deployment to edge nodes in response to client requests, leading to increased latency due to distant application access, which is particularly problematic for applications like mobile map services and AI control systems.
Innovation Solution
An application management system that includes a deployment controller, orchestration manager, and DNS service, which identifies the need for application deployment to geographically closer edge nodes upon client requests, using Kubernetes server-client provisioning architecture to containerize applications and deploy them on eligible edge nodes, thereby reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If applications are deployed to distant cloud data centers, then resource sharing and scalability are improved, but communication latency increases for latency-sensitive applications
Solution Approach 1:
The patent segments the monolithic cloud data center into distributed edge nodes deployed geographically closer to end users. Each edge node provides localized application deployment while maintaining connection to the central orchestration system, enabling resources to be shared across multiple locations and reducing communication latency for local users.
Solution Approach 2:
The patent adds a spatial dimension to cloud deployment by distributing applications across multiple geographic locations (edge nodes) rather than concentrating them in a single data center. This dimensional expansion allows users to access applications from the nearest edge node, reducing latency while maintaining resource sharing capabilities.
2Loss of time
If applications are pre-deployed to multiple edge nodes, then access latency is reduced, but resource waste occurs when applications are not needed
Solution Approach 1:
The patent implements preliminary action by pre-positioning application containers at edge nodes in a dormant state, ready for rapid deployment when requested. The orchestration system maintains metadata about available applications at each edge node, enabling fast local deployment without continuous resource consumption for unused applications.
Solution Approach 2:
The patent applies dynamics by making edge node resource allocation adaptive and on-demand. Applications are deployed to edge nodes dynamically based on real-time user requests and local conditions, allowing the system to optimize between latency reduction and resource efficiency by activating only the necessary instances.
3Adaptability or versatility
If user self-provisioning is implemented, then service flexibility is improved, but deployment complexity increases for on-demand edge application provisioning
Solution Approach 1:
The patent introduces an intermediary orchestration system that mediates between users and edge nodes. This intermediary handles the complex tasks of application packaging, containerization, and deployment coordination, allowing users to request applications through simple interfaces while the orchestration system manages the technical complexity of edge node provisioning and deployment.
Data Source
AI summary
A method is provided method to control deployment of an application over a network in response to a client request sent over the network to access the application comprising: capturing at one or more first computing machines coupled to the network, an identifier of the requested application from the client request; sending information over the network from the one or more first computing machines coupled to the network to one or more second machines coupled to the network, wherein the information identifies the requested application and identifies a network address of an edge node at which to deploy the requested application; receiving the information at the one or more second machines coupled to the network; and causing by the one or more second machines coupled to the network, deployment of the application over the network to the edge node at the identified network address, based at least in part upon the received information.