Virtual Node Management for Stable Mobile Cluster Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cluster computing technologies face challenges in managing nodes with fluctuating resource demands due to mobility, network instability, and battery limitations, particularly in mobile nodes, leading to inefficient resource utilization and stability issues.
Innovation Solution
A method involving a virtual node that receives registration requests from nodes, performs registration based on node information, identifies suitable nodes for pod transmission, and manages resource allocation using a control plane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If mobile nodes are added to ensure resource mobility, then resource flexibility is improved, but network stability deteriorates due to fluctuating network states
Solution Approach 1:
The patent introduces a virtual node as an intermediary layer between the control plane and physical nodes. This virtual node receives registration requests from physical nodes, manages their registration information, and handles pod distribution. By inserting this intermediary, the system can maintain stable control over mobile nodes without being directly affected by their network fluctuations, thus resolving the contradiction between resource mobility and network stability.
2Adaptability or versatility
If nodes with fluctuating resources are allowed in the cluster, then resource utilization flexibility is improved, but cluster configuration reliability deteriorates
Solution Approach 1:
The patent creates a virtual copy of the node management interface through the virtual node. Physical nodes with fluctuating resources register with the virtual node, which maintains a stable virtual representation of these nodes. The control plane interacts with this stable virtual representation rather than directly with the fluctuating physical nodes, allowing flexible resource utilization while maintaining configuration reliability.
3Device complexity
If direct node management by control plane is used, then management simplicity is maintained, but scalability deteriorates when managing large numbers of nodes
Solution Approach 1:
The patent segments the node management function by separating it into two parts: the virtual node that handles registration and information management, and the control plane that handles pod distribution and high-level control. This segmentation allows the system to scale to large numbers of nodes by distributing registration requests to the virtual node without overwhelming the control plane, thus improving scalability while maintaining management simplicity.
Data Source
AI summary
A method of performing communication by a virtual node includes receiving a registration request from at least one node of a plurality of nodes, performing registration of the at least one node, based on information related to the at least one node, receiving, from a control plane, a pod indicating at least one distributable computing unit of a task to be completed, identifying a first node from among the at least one node to which the pod is to be transmitted, based on the information related to the at least one node, and transmitting the pod to the first node. The plurality of nodes and the virtual node are controlled by the control plane.