Behind the scenes of the async runtime, various things are taking place to efficiently execute the Futures provided by the user. By observing such runtime-level events, you can often gain deeper insights into your program, which can be helpful for troubleshooting and more.

Unfortunately, as of now, there are no monitoring tools available that can visualize the events occurring within such an async runtime.

Therefore, I'm currently exploring whether we can integrate tracing into Tokio itself to visualize runtime events.

In this session, I'd like to share the insights I've gained through this attempt. Specifically, I plan to cover the following topics:

• Existing tools for monitoring async applications and scenarios that are difficult to cover with them
• Motivation for runtime tracing
• Similar existing technologies (e.g., Go trace, JFR)
• What kind of runtime information can be included in tracing?
• How can we capture stack traces of async tasks?
• Methods for visualizing the trace