Intelligent Numerical Simulation Operation Management and Feedback Method

Abstract

The invention discloses an intelligent numerical simulation job management and feedback method. The method comprises the following steps that a calculation node continuously samples node state information; a server program excludes abnormal nodes, arranges nodes according to priorities of the nodes and combines the nodes into calculation resources waiting to be dispatched; a user uses a client program to send a numerical simulation job calculation request; the server program preferentially assigns the job calculation request to the calculation node with high priority for calculation under the condition of sufficient calculation resources; after the job calculation is ended, the client program analyzes a job log to obtain a job completion state; and the client program performs feedback according to the job completion state and updates the priority of the calculation node. According to the method, a job can be effectively prevented from being repeatedly transmitted to the abnormal nodes for calculation, and the priorities of the nodes with high calculation efficiency are improved gradually, so that the calculation success rate and efficiency of the whole numerical simulation job are increased and improved, and the working efficiency of scientific and research personnel is effectively improved; and the method has relatively high universality.

Description

technical field [0001] The invention relates to the technical field of numerical simulation job scheduling management computing resource optimization, in particular to a numerical simulation job management computing resource intelligent filtering and feedback optimization method. Background technique [0002] The back-end calculation resource optimization technology of numerical simulation operation calculation is one of the key technologies for the effective development of reservoir numerical simulation work in the western South China Sea basin. [0003] Most of the existing numerical simulation job management technologies are node scheduling management with fixed priority, or node random scheduling management, but computing nodes are affected by many uncontrollable factors, and various factors may lead to the existence of bad nodes. As a result, the overall calculation success rate of the numerical simulation job is reduced, and the node random scheduling management can re...

AI Technical Summary

Problems solved by technology

[0003] Most of the existing numerical simulation job management technologies are node scheduling management with fixed priority, or node random scheduling management, but computing nodes are affected by many uncontrollable factors, and various factors may lead to the existence of bad nodes. As a result, the overall calculation success rate of the numerical simulation job is reduced, and the node random scheduling management can reduce the impact of bad nodes on the whole, but it cannot effectively improve the calculation efficiency

Reservoir numerical simulation users have repeatedly reported that operation calculations fail irregularly, which is actually a problem of the decline in the quality of back-end computing resources, which greatly reduces the efficiency of scientific research work. Therefore, there is an urgent need for a Job Management Scheduling Method for End Computing Resource Quality

Images