Concurrent Computing Exception Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Concurrent computing systems, such as those using MATLAB, face challenges in efficiently handling large and complex computations that require significant resources, as existing methods like MPI often lead to abrupt termination of processes upon errors, disrupting communication channels and necessitating re-establishment, which is undesirable.
Innovation Solution
A method is introduced to initiate and manage a group of instances in a concurrent computing process, establishing communication channels, detecting interrupt requests, and terminating computations while maintaining communication, allowing for controlled termination and reinitialization of processes without disrupting the entire system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional MPI methods are used to establish communication channels between multiple workstations, then concurrent computing capability is enabled, but upon error the processes are abruptly terminated and communication channels are disconnected requiring re-establishment
Solution Approach 1:
The patent implements exception handling mechanisms that anticipate and prepare for errors before they cause system failure. By establishing try-catch blocks and exception handlers at the beginning of parallel computing operations, the system creates a protective framework that prevents abrupt terminations and maintains communication channels even when errors occur during computation.
Solution Approach 2:
The patent transforms the harmful effect of errors and exceptions into a beneficial outcome by implementing exception handling that converts potential system failures into controlled, non-disruptive events. When exceptions occur, the handlers catch them and allow the parallel computing process to continue, effectively turning errors into opportunities to demonstrate system robustness and maintain operational continuity.
2Power
If multiple workstations are used to perform concurrent computing, then computing power is increased to handle large and complex problems, but system complexity increases requiring coordination and communication management
Solution Approach 1:
The patent introduces exception handling mechanisms as intermediary components between the parallel computing processes and the communication channels. These handlers act as mediators that manage errors and exceptions, preventing them from disrupting the coordination between multiple workstations. The exception handlers absorb and manage complexity by providing a standardized interface for error management across all participating processes.
Solution Approach 2:
The patent implements universal exception handling mechanisms that can manage multiple types of errors and exceptions across different workstations and communication protocols. The exception handlers are designed to be multi-functional, capable of catching various exception types, managing different communication channel failures, and maintaining system operation under diverse error conditions, thereby reducing overall system complexity through standardized management.
3Reliability
If communication channels are maintained during error conditions, then system reliability is improved, but resource management becomes more complex requiring controlled termination mechanisms
Solution Approach 1:
The patent implements dynamic resource management through exception handling that adapts to error conditions in real-time. The exception handlers dynamically adjust resource allocation and process termination based on the specific error encountered, allowing communication channels to be maintained when appropriate while enabling controlled termination when necessary. This dynamic approach simplifies resource management by providing automated, context-aware decision-making rather than requiring static, manual resource management configurations.
Data Source
AI summary
A system initiates multiple instances of a concurrent computing process, establishes a communication channel among the multiple instances, initiates execution of a computational job on the multiple instances, detects an interrupt request on one of the multiple instances, and terminates execution of the computational job while maintaining communication among the multiple instances via the communication channel.