Modern IT operations demand proactive issue detection, and observability has become crucial for understanding today’s complex, distributed application landscapes – including mainframe environments. This session introduces the fundamentals of observability and highlights its value, focusing on OpenTelemetry as the emerging industry standard for capturing metrics, logs, and traces.
We’ll explore how OpenTelemetry can be applied to Linux on IBM Z, demonstrating how infrastructure-level telemetry can be extended down to the hardware layer through instrumentation of the IBM Z Hardware Management Console (HMC). Attendees will learn how capacity metrics, LPAR activity, and hardware events can be integrated into enterprise observability pipelines, facilitating correlation between application behavior and the underlying system infrastructure.
The session also examines zero-code instrumentation techniques for Linux on IBM Z, enabling existing workloads to emit standardized telemetry without application source code modifications. We’ll showcase how these capabilities can be implemented with minimal operational disruption. Increasingly, this includes the observability of AI and LLM-powered workloads, allowing us to monitor model performance, prompt behavior, and overall system impact.
Finally, we’ll provide an overview of current and planned activities within the OpenTelemetry community and the Open Mainframe Project, focused on advancing consistent, enterprise-wide observability across IBM Z and LinuxONE – from hardware to applications.