EANA Module Failure Detection in Distributed Applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current failure detection systems in distributed application environments are application-specific and do not guarantee consistent failure detection, especially when failures occur in hardware, and different applications use varying timeouts, complicating unified management.
Innovation Solution
A method and system utilizing an Enterprise Application Nexus API (EANA) that sends status messages to application nodes, detects failures, and coordinates lock and unlock messages across a tree structure of applications to suspend or resume operations, ensuring comprehensive failure management including hardware failures, through an EANA module and proxy that handles aware and unaware application nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exception handling is used for failure detection, then application-specific errors can be detected, but hardware failures cannot be reliably detected and the system lacks unified timeout management
Solution Approach 1:
The patent introduces an intermediary component (failure detection system with centralized timeout management) that mediates between individual application nodes and the distributed system. This intermediary receives status messages from all nodes, uniformly manages timeout detection across the entire system, and coordinates failure responses, thereby achieving both reliable failure detection and unified timeout management.
Solution Approach 2:
The failure detection system is designed with universal functionality that serves multiple purposes: it detects application-specific exceptions, detects hardware failures through heartbeat monitoring, manages timeouts uniformly across all nodes, and coordinates failure responses. This multi-functional design eliminates the need for separate detection mechanisms for different failure types.
2Reliability
If application-specific exception handling is implemented, then individual application failures can be caught, but system-wide consistent failure management is achieved
Solution Approach 1:
The patent introduces an intermediary component (failure detection system with centralized timeout management) that mediates between individual application nodes and the distributed system. This intermediary receives status messages from all nodes, uniformly manages timeout detection across the entire system, and coordinates failure responses, thereby achieving both reliable failure detection and unified timeout management.
Solution Approach 2:
The patent merges multiple failure detection approaches (exception handling, heartbeat monitoring, timeout detection) into a single unified failure detection system. This consolidation reduces the number of separate components and simplifies the overall system architecture while maintaining comprehensive failure detection capabilities.
3Adaptability or versatility
If different applications use various timeouts, then each application can optimize for its specific needs, but unified timeout management and consistent failure detection are achieved
Solution Approach 1:
The patent applies local quality by allowing each application node to have its own timeout configuration and status message format tailored to its specific requirements, while the centralized failure detection system uniformly processes all these diverse inputs using consistent detection logic, thereby achieving both application-specific optimization and system-wide consistency.
4Reliability
If a centralized failure detection system is implemented, then unified timeout management and consistent failure detection are achieved, but the system complexity and coordination overhead increase
Solution Approach 1:
The patent extracts the complex failure detection and timeout management logic from individual application nodes and concentrates it in a separate centralized failure detection system. This extraction simplifies the application nodes themselves while isolating the complexity into a dedicated component that can be independently managed and maintained.
Data Source
AI summary
A mechanism for managing failure of applications in a distributed environment is disclosed. A method includes detecting failure in an application node among a plurality of application nodes when the application node does not respond to a status message. The method further includes routing, by enterprise application nexus application processing interface (EANA) module, a first lock message to the failed application node and ancestors' nodes of the failed application node when the failed application node is an aware application node.