How Go 1.20/1.21 Revamps Runtime Tracing for Faster Debugging

Introduction: Go's runtime/trace package provides a powerful way to understand and debug Go programs, with visualisation via go tool trace or tools like gotraceui.

Problem: Historically, execution tracing suffered from high overhead, limited scalability, uncertainty about when to start tracing to capture specific issues, and difficulty associating trace data with source code.

Improvements in Go 1.20 and Go 1.21: a more efficient tracing mechanism that lowers overhead and improves scalability; a new API that makes starting, stopping, and scoping traces easier; enhanced tools for visualising and exploring trace data; and new capabilities to link trace data directly to program source code.

Result: These enhancements turn Go execution tracing into a stronger, more user‑friendly tool, helping developers more easily identify and resolve concurrency bottlenecks.

Example: Using the new API in Go 1.21 to start a one‑second trace, perform work, and stop the trace, then visualising the result with go tool trace main.trace:

func main() {
  // Start a trace lasting for 1 second.
  trace.Start(trace.Options{Duration: time.Second})

  // Do some work.
  for i := 0; i < 1000000; i++ {
    // ...
  }

  // Stop the trace.
  trace.Stop()
}

Conclusion: With the recent improvements, Go's execution tracing becomes an essential part of a Go developer’s toolbox for diagnosing and fixing concurrency problems.