Java & JVMConference - Short25min
Who's Slowing Down Your Application? Finding the Culprit When Logs and Metrics Fail
The session demonstrates how the open-source Java Flight Recorder (JFR) transforms performance debugging by revealing detailed runtime insights. It showcases real production examples where JFR quickly identified hidden issues, significantly reducing troubleshooting time and stress, and advocates adopting JFR as a standard tool for continuous performance monitoring.
talk.summaryAiDisclaimer
Michał GniadekTen Square Games
talkDetail.whenAndWhere
Wednesday, June 17, 09:00-09:25
Room 3
talks.roomOccupancytalks.noOccupancyInfo
When things in production aren’t running as fast as they should, and your logs and metrics are coming up empty, a real investigation begins. In this session, I’ll show you how the free, open-source Java Flight Recorder (JFR) can completely flip the script, dramatically accelerating the way you debug performance issues. Instead of hunting for the culprit blindly, JFR allows us to look "under the hood" of a running application and pinpoint the source of our problems with surgical precision.
I will share how we used JFR to solve real, non-obvious production mysteries, saving us a massive amount of time and stress in the process. My goal is for JFR to become a permanent resident in your production environments, providing you with invaluable insights into your application's behavior :)
I will share how we used JFR to solve real, non-obvious production mysteries, saving us a massive amount of time and stress in the process. My goal is for JFR to become a permanent resident in your production environments, providing you with invaluable insights into your application's behavior :)
Michał Gniadek
It all started with Minecraft and my first Java mods. Today, I use those same tools to manage high-volume global traffic. I’m deeply fascinated by Java performance and its internal organs, which I enjoy dissecting right up until I hit C code. On a daily basis, I focus on high-performance system optimization and hunting down non-trivial production bottlenecks