Hierarchical Network Liveness Detection for Large Managed Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Detecting the liveness of managed nodes across various topologies, especially in corporate networks, is challenging due to the difficulty in reaching all nodes and the high resource requirements, particularly when the number of nodes is large.
Innovation Solution
A network management server creates clusters based on node coordinates, designating a coordinator and leader node to track liveness, reducing direct communication to the server by using a hierarchical structure with heartbeat messages between nodes and consolidating liveness information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the network management server directly detects liveness of each managed node, then the liveness detection is straightforward, but the server becomes overloaded and resource consumption becomes prohibitively costly when the number of nodes is large
Solution Approach 1:
The patent segments the managed network into hierarchical clusters with coordinator nodes and leader nodes. Each coordinator manages a subset of nodes, and leaders manage groups of coordinators. This segmentation distributes the liveness detection burden across multiple nodes rather than concentrating it all at the server, resolving the contradiction between reliable detection and server resource consumption.
Solution Approach 2:
The patent introduces intermediary nodes (coordinators and leaders) between the server and managed nodes. These intermediaries consolidate liveness information from multiple nodes and report to higher-level nodes or the server, acting as mediators that reduce the direct communication burden on the server while maintaining detection reliability.
2Loss of information
If the server communicates with all managed nodes directly, then complete node status information is obtained, but the number of sockets, memory, and connections required becomes prohibitively costly
Solution Approach 1:
The patent merges liveness information from multiple managed nodes at coordinator nodes, and then merges coordinator status information at leader nodes. This hierarchical merging consolidates numerous individual node status reports into aggregated cluster status reports, maintaining information completeness while dramatically reducing the number of connections required at the server level.
Solution Approach 2:
The patent introduces a hierarchical dimension to the communication structure, organizing nodes into multiple levels (managed nodes → coordinators → leaders → server). This dimensional change transforms the flat many-to-one communication model into a multi-level hierarchical model, reducing the connection burden while preserving information flow.
3Adaptability or versatility
If mobile devices are tracked across different locations in the corporate network, then comprehensive device monitoring is achieved, but detecting server-initiated liveness becomes difficult due to topology changes
Solution Approach 1:
The patent implements a dynamic hierarchical structure where coordinator and leader nodes are selected based on current network conditions and node locations. As mobile devices move between locations, the hierarchy dynamically reconfigures to maintain effective liveness detection, adapting to topology changes while preserving monitoring capability.
Solution Approach 2:
The patent inverts the traditional client-server liveness detection model by having managed nodes actively report to coordinators, who then report to leaders, and finally to the server. This inversion transforms the problem from server-initiated detection (which fails with mobile nodes) to node-initiated reporting, solving the liveness detection difficulty while maintaining adaptability to location changes.
Data Source
AI summary
A request is received by a network management server, from a managed node, to get cluster information. The cluster information identifies a coordinator node and a leader node (a node hierarchy) that are used to track liveness of the managed node. The coordinator node and the leader node may be identified based on being in the same location as the managed node. The cluster information is sent to the managed node to make the managed node aware of the hierarchy. The coordinator node consolidates liveness of the nodes in its grouping in the cluster. The coordinator node sends a first liveness message of the managed node to the leader node. The leader consolidates a group of coordinator nodes by sending a second liveness message of the managed node to the network management server. This gives the network management server a status of all the managed nodes in the cluster.