Method for predicting operation duration of batch processing task and related device

Abstract

The invention discloses a batch processing task running time prediction method and a related device, and the method comprises the steps: taking a batch node in a batch processing task as a target batch node, obtaining a job set executed by the target batch node, taking one job in the job set as a target job, and obtaining the target job; and obtaining the running duration of the target job according to the daytime online transaction volume, the I / O throughput, the CPU resource and the CF resource of the target job, obtaining the running duration of the target batch node according to the running duration of the target job, and obtaining the running duration of the batch processing task according to the running duration of the target batch node. Therefore, the operation duration of the job can be accurately predicted through the factors influencing the operation duration, including the daytime online transaction volume, the I / O throughput, the CPU resources and the CF resources, so that the operation duration of the batch nodes and the operation duration of the batch processing tasks are obtained, the operation monitoring of the batch processing tasks is realized, and the operation efficiency of the batch processing tasks is improved. Therefore, resource conflicts can be avoided in advance by increasing resources, pausing execution of some jobs and the like.

Description

technical field [0001] The present invention relates to the technical field of data processing, in particular to a method for predicting the running time of a batch processing task and a related device. Background technique [0002] During the operation and maintenance process of the mainframe system, the core system will perform batch processing through scheduled execution and batch execution when the business volume is small, such as the time period from 2:00 am to 4:00 am. A batch task consists of multiple batch nodes. A batch node is a complete set of job flow. A job flow includes multiple jobs. Multiple jobs may be executed serially or in parallel. Therefore, the running time of different batch nodes and even batch tasks may be different. [0003] If the running time of the current batch node is too long, it may affect the operation of the next batch node, it may affect the changes that need to be avoided during the batch execution period, and even affect the smooth op...

AI Technical Summary

Problems solved by technology

[0004] In related technologies, the running time of batch processing tasks is predicted by human experience, but this method is too dependent on human experience, the prediction accuracy is not high, different experiences may lead to inconsistent prediction results, and the cost of human consumption is high

Images