The pattern shows you how to analyze batch jobs on a mainframe with Watson Machine Learning for z/OS. This includes ingesting batch job run time data from SMF Type 30 records, exploring periodicity and correlation, predicting elapsed time, and detecting anomalies.
Running batch jobs is critical to the effective operation of a mainframe. Every day, a typical mainframe might have 10,000 – 60,000 jobs running day and night. And on days that fall at the beginning or end of a month, quarter, or year, workloads can be as much as twice that size. Effective batch processing is critical to maintaining the high productivity demands of many businesses today.
You can use IBM® Watson™ Machine Learning for z/OS to analyze your batch processing and extract insights to enhance the following aspects of your batch operation:
- Trend and seasonality of batch jobs’ elapsed time according to workload change
- Impact of transactions and other business operations on the elapsed time of batch jobs
- Prediction of elapsed time for long-running jobs
- Identification of potentially abnormal job instances and transaction volumes
- You can work with Watson Machine Learning for z/OS using a web browser.
- Watson Machine Learning for z/OS provides Jupyter Notebooks for you to code in Python and R.
- Watson Machine Learning for z/OS provides Modeler Flow for you to explore data and train model in canvas by drop and down.
- You can read mainframe native files such as SMF Type 30 record with Python notebooks based on the Mainframe Data Service that’s included in Watson Machine Learning for z/OS.
Ready to get started? Check out the README.md for detailed instructions on how to:
- Download the .zip file (for Windows or Mac) or .tar.gz file (for Linux) in the “ProjectZIP” folder to your computer.
- Login to Watson Machine Learning for z/OS with your username and password.
- Click on “New project” under “Getting started.”
- Select the “From File” tab, browse in your computer and select the file you downloaded in step 1, then click “OK.”
- Confirm that the new project has been added. It should include 4 notebooks, 2 flows, and 5 data sets.