Distributed Network Floating IP Address Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for searching floating IP addresses in distributed networks, especially those with mobile nodes and unstable internet connections, face inefficiencies due to high traffic loads and the need for frequent server replacements, and are not scalable with increasing numbers of nodes and service addresses.
Innovation Solution
A method utilizing peer-to-peer communication between representative nodes through a network ring and address range, allowing for efficient searching and registration of floating IP addresses without relying on broadcasting, enabling stable network construction even with mobile nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If broadcasting scheme is used to search for floating IP address, then all nodes can be reached, but traffic load increases significantly
Solution Approach 1:
The network is segmented into multiple sub-networks with representative nodes. Instead of broadcasting to all nodes globally, the search is segmented into local sub-network searches coordinated through representative nodes, reducing overall traffic load while maintaining discovery completeness.
Solution Approach 2:
Representative nodes act as intermediaries between the querying node and the target node. The representative nodes receive and forward search requests, eliminating the need for direct broadcasting to all nodes and significantly reducing network traffic.
2Speed
If local server is used to store IP information, then search operation is rapid, but server performance becomes bottleneck as data capacity increases
Solution Approach 1:
The centralized server is segmented into multiple distributed representative nodes. Each representative node stores and manages a portion of the IP information locally, enabling parallel search operations that maintain speed while scaling capacity across multiple nodes.
Solution Approach 2:
Each representative node independently stores and manages its own IP information, performing local search operations without requiring centralized server processing. This self-service approach eliminates the server bottleneck and enables horizontal scaling.
3Adaptability or versatility
If server capacity is increased to handle more data, then more nodes can be supported, but it is inefficient to replace or add servers
Solution Approach 1:
The system uses self-organizing representative nodes that automatically join and integrate into the distributed network. New nodes can be added without manual server configuration or replacement, enabling scalable network growth while maintaining simple management operations.
Solution Approach 2:
The distributed network structure dynamically adapts to accommodate new nodes and representative nodes. The system automatically reconfigures search paths and routing when nodes are added or removed, providing scalability without requiring static server infrastructure planning.
4Loss of energy
If distributed processing is used, then traffic load is reduced, but data must be transmitted to all nodes including non-desired services
Solution Approach 1:
Representative nodes serve as intermediaries that filter and forward data selectively. They receive search requests and forward them only to relevant sub-networks or nodes, preventing unnecessary data transmission to nodes that do not provide desired services while maintaining distributed processing benefits.
Data Source
AI summary
A method of registering a new node in a network and a method of searching for a floating Internet Protocol (IP) address through a distributed network are provided. The method of searching for the floating IP address may include receiving, by a representative node of a head network ring, from an external node, a message comprising an Identification (ID) of a target node, determining whether the ID of the target node is registered in the head network ring, based on a ring information table, and transferring the message to a representative node of a child ring having an address range of the target node among child rings, when the ID of the target node is not registered in the head network ring.