Browse Prior Art Database

A Method to Discover Job Flow from Job Execution Log

IP.com Disclosure Number: IPCOM000199389D
Publication Date: 2010-Sep-01
Document File: 1 page(s) / 34K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a system for discovering implicit job flow from a job execution log.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

A Method to Discover Job Flow from Job Execution Log

Disclosed is a system for discovering implicit job flow from a job execution log.

A computer system such as mainframe computer executes multiple jobs in a job flow. Some of job flows are explicitly defined in either JCL or job net which is defined with a

job scheduler. However, there are job flows which are implicitly

ob

A and job B are started with timers independently, but job B uses output from job A, hence job A always needs to be started prior to job B.

The disclosed system extracts such implicit job flows with combining the following two information.
- static dependency graph among jobs based on relationship between a job and a dataset - a job execution log

Two jobs have relationship when they access the same dataset. Some existing asset anlaysis tools can discover such relationship as the result of analysis of JCL or program source code. However, as a program often access a dataset in both input and output, it is unable to determine which job accesses the dataset first. In other words, the dependency graph can be bi-directional, hence the sequence of flow in the graph cannot be deterministic.

In order to identify the job flow (sequence of jobs), the dependency graph needs to be changed to uni-directional graph.

defined. For instance,

j

1