Offloading Health-Check Policy to Nodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud computing systems experience slower failure detection due to network resource consumption from periodic polling for node state information, leading to higher Mean Time To Recovery (MTTR) and lower availability.
Innovation Solution
Offloading health-checking policy to nodes, where each node receives a failure policy and autonomously reports faults to the cloud health monitor when predefined conditions are met, reducing network usage and enabling faster metric gathering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If centralized polling is used to monitor node health, then management control is maintained, but network resource consumption increases and failure detection speed decreases
Solution Approach 1:
The patent implements self-service by enabling nodes to autonomously monitor their own health status and automatically report failures when thresholds are exceeded. Each node executes local health-checking policies, compares metrics against predefined thresholds, and autonomously generates failure reports without requiring centralized polling, thus reducing network consumption while improving detection speed
Solution Approach 2:
The patent inverts the traditional polling architecture by switching from centralized management-initiated monitoring to distributed node-initiated reporting. Instead of the management component continuously querying nodes, nodes continuously self-monitor and only contact the management component when failures occur, fundamentally reversing the communication initiation pattern to eliminate unnecessary network traffic
2Reliability
If centralized polling is used for health monitoring, then comprehensive node monitoring is achieved, but Mean Time To Recovery (MTTR) increases
Solution Approach 1:
The patent applies preliminary action by pre-configuring health-checking policies and failure thresholds on each node before deployment. Nodes continuously evaluate their health status against these pre-established criteria, enabling immediate failure detection and reporting without waiting for polling cycles, thus reducing MTTR while maintaining comprehensive monitoring coverage
Solution Approach 2:
The patent ensures continuity of useful action by implementing continuous local health monitoring on each node. Rather than periodic centralized polling, nodes continuously track their own metrics and maintain readiness to report failures instantly, ensuring uninterrupted health monitoring coverage across the distributed system
3Productivity
If periodic polling is implemented for state information collection, then centralized control is maintained, but system responsiveness decreases
Solution Approach 1:
The patent applies segmentation by dividing the centralized monitoring function into distributed autonomous units. Each node independently executes health-checking policies and maintains local state information, segmenting the monolithic polling architecture into independent, self-managing units that respond immediately to local conditions without centralized coordination overhead
Data Source
AI summary
Methods and systems for offloading health-checking policy in a distributed management environment are provided. A failure policy is received at a node of a cloud from a cloud health monitor. The node transmits a notification to a health monitor of the node that the node has failed when the failure policy is satisfied. The node reports at least one fault based on the satisfied failure policy to the cloud health monitor.