zManav: Db2 for z/OS Wellness Monitor
Ravichandran, Manigandan (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024120933921
https://urn.fi/URN:NBN:fi:amk-2024120933921
Tiivistelmä
The zManav, Db2 for z/OS Wellness Monitor, is an AI-driven solution that automates the monitoring and analysis of Mainframe Db2 for z/OS operational logs. The development research addresses the challenge of analyzing complex mainframe syslogs, which are prone to anomalies during high workload periods, disaster recovery, during major mainframe infrastructure h/w and s/w upgrade and maintenance activities.
By leveraging periodic automated reporting from Db2 for z/OS MSTR JESMSLG integrated with graphical visualizations, the zManav streamlined log analysis and enhanced administrative efficiency.
Using Design Science Research method, the author developed an automated syslog parsing and anomaly detection tool-zManav which tailored to Db2 MSTR JESMSGLG logs. The Design Science research method is categorised into five phases.
They are as follows:
(i)Problem and Motivation: Transform humanistic monitoring to AI based monitoring and alerting.
(ii)Objective Definition: Develop a technique to transform logs to structural format for data analytics.
(iii)Design and Development: Using pandas python libraries in Databricks, produce metrics and alerts.
(iv)Demonstration: Demonstrate the metrics in time-series analysis framework using zManav tool.
(v)Communication: Document the research development as Master’s thesis and training to the team.
The AI-driven analysis and monitoring process involved extracting relevant data from Mainframe Db2 system logs and moved to Databricks as text files for Dat analytics using Pandas Data Frame and the processed data is then stored as TSV (Tab Separated Value) format. The resulting metrics using Data analytics, facilitated behaviour analysis by identifying trends associated with selected message IDs and highlighting new message IDs that might require attention.
The zManav not only streamlined the monitoring and analysis of mainframe Db2 logs but also established a foundation for smarter, data-driven mainframe management. By automating routine checks, enhancing predictive capabilities, and fostering proactive responses, the tool demonstrated its potential to transform mainframe administration practices and improve overall operational efficiency.
By leveraging periodic automated reporting from Db2 for z/OS MSTR JESMSLG integrated with graphical visualizations, the zManav streamlined log analysis and enhanced administrative efficiency.
Using Design Science Research method, the author developed an automated syslog parsing and anomaly detection tool-zManav which tailored to Db2 MSTR JESMSGLG logs. The Design Science research method is categorised into five phases.
They are as follows:
(i)Problem and Motivation: Transform humanistic monitoring to AI based monitoring and alerting.
(ii)Objective Definition: Develop a technique to transform logs to structural format for data analytics.
(iii)Design and Development: Using pandas python libraries in Databricks, produce metrics and alerts.
(iv)Demonstration: Demonstrate the metrics in time-series analysis framework using zManav tool.
(v)Communication: Document the research development as Master’s thesis and training to the team.
The AI-driven analysis and monitoring process involved extracting relevant data from Mainframe Db2 system logs and moved to Databricks as text files for Dat analytics using Pandas Data Frame and the processed data is then stored as TSV (Tab Separated Value) format. The resulting metrics using Data analytics, facilitated behaviour analysis by identifying trends associated with selected message IDs and highlighting new message IDs that might require attention.
The zManav not only streamlined the monitoring and analysis of mainframe Db2 logs but also established a foundation for smarter, data-driven mainframe management. By automating routine checks, enhancing predictive capabilities, and fostering proactive responses, the tool demonstrated its potential to transform mainframe administration practices and improve overall operational efficiency.