System Insights Agent Demo: Late high CPU job analysis scenario

In this scenario, we used natural language interactions with IBM Z OMEGAMON Insights and IBM Z Workload Scheduler Insights agents within IBM watsonx Assistant for Z to troubleshoot two interconnected issues.

The root cause was a long-running job that triggered high CPU usage on an LPAR. Eventually, the job abended, preventing a dependent batch job—responsible for generating a critical report—from starting.

By combining real-time monitoring data and scheduling insights through an AI-driven conversation, we quickly identified the issue and resolved it by manually updating the job status. This allowed the dependent job to execute successfully and ensured the report was produced as expected.