Network Subgraph Aggregation for Server Communication Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques are inefficient in processing and managing communication information between thousands of servers in an enterprise, leading to delays in implementing administrative domain-wide policies.
Innovation Solution
A method and system that utilize network subgraphs to efficiently manage server communications by generating unit network subgraphs for each server, updating edges based on policy expressions, and evaluating administrative domain-wide policies to perform specified actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional techniques are used to process communication information between servers, then the system can handle all communications, but the processing efficiency is low and delays occur
Solution Approach 1:
The patent segments the large-scale network communication data into unit network subgraphs, each representing a specific server's communication patterns. This segmentation allows parallel processing of multiple servers' data independently, significantly improving processing efficiency and reducing delays compared to handling all communications as a single monolithic dataset.
Solution Approach 2:
The patent transforms communication data from a flat, centralized structure into a multi-dimensional graph structure with nodes representing servers and edges representing communications. This dimensional transformation enables efficient querying and analysis by allowing policy expressions to traverse and evaluate relationships across the network topology, rather than processing linear sequences of communication records.
2Reliability
If all communication information between thousands of servers is processed centrally, then complete policy enforcement is achieved, but system complexity and processing burden increase
Solution Approach 1:
The patent divides the centralized policy enforcement system into distributed unit network subgraphs for each server. Each subgraph independently evaluates policies relevant to its associated server, ensuring complete policy enforcement while distributing the computational burden across multiple independent processing units rather than overwhelming a single central system.
Solution Approach 2:
The patent pre-computes and stores unit network subgraphs for each server, capturing their communication patterns in advance. This preliminary structuring of data allows policy expressions to be evaluated efficiently against pre-organized network relationships, reducing the complexity of real-time policy enforcement while maintaining completeness.
Data Source
AI summary
Communications between servers of a network are received and processed to generate network subgraphs. The network subgraph may be used for implementing administrative domain wide policies, for presenting information on a user interface, or for other purposes. A unit network subgraph is generated for each server. The unit network subgraph represents communications of the server with other servers. A network subgraph is generated by aggregating information of unit network subgraphs servers included in the network subgraph. The network subgraph may be used to evaluate an expression specified by an administrative domain wide policy. The result of evaluation of the expression may be used to determine whether actions specified by the policy are executed. The network subgraph may be used for processing network flow queries, for example, for displaying network information on a user interface.