Profile Your Production Java* Workload in the Cloud

Profiling Java* workloads for the cloud can be challenging. The common reasons are because the workloads:

• Are generally long-running services
• Are launched by a scheduler or orchestrator using a headless user account
• Must be profiled in a production environment because it’s hard to create a comparable test environment

This session shows how Intel® VTune™ Profiler can remove these obstacles, whether your applications run in public, private, or hybrid cloud environments.

This session covers how to:

• Set up Intel VTune Profiler for remote cloud profiling (in other words, an Intel VTune Profiler server versus a remote collection).
• Profile your running Java process with minimal overhead.
• Review Intel VTune Profiler results to understand where performance can be improved, and to find and fix bottlenecks.

As a bonus, see a demonstration of the Flame Graph feature in Intel VTune Profiler that detects hot spots and hot code paths in Java workloads.

To get the most from this session, you must have a basic understanding of the Java Development Kit (JDK), the Java Virtual Machine (JVM), and how configurations can impact performance.

Get the stand-alone Intel VTune Profiler or as part of the Intel® oneAPI Base Toolkit— a core set of tools and libraries for developing high-performance, data-centric applications across diverse architectures.