There's little doubt that Go provides some of the best ancillary tools available to any compiled language out there. Compiling to native code on a myriad of systems, deadlock detection, pprof, fmt, and more allow you to not just build high-performance applications, but also test them and format them.
This hasn't stopped the community from developing other tools that can be used for debugging or aiding your concurrent and/or distributed code. We'll take a look at a few great tools that may prove worthy of inclusion in your app, particularly if it's highly visible or performance critical.
As you are probably now well aware, Go's pprof is extremely powerful and useful, if not exactly user-friendly.
If you love pprof already, or even if you find it arduous and confusing, you may love Nitro profiler twice as much. Coming from Steve Francia of spf13, Nitro profiler allows you to produce even cleaner analyses of your application and its functions and steps, as well as providing more usable a/b tests of alternate functions.
Tip
Read more about Nitro profiler at http://spf13.com/project/nitro.
You can get it via github.com/spf13/nitro.
As with pprof, Nitro automatically injects flags into your application, and you'll see them in the results themselves.
Unlike pprof, your application does not need to be compiled to get profile analysis from it. Instead, you can simply append -stepAnalysis to the go run command.
Heka is a data pipeline tool that can be used to gather, analyze, and distribute raw data. Available from Mozilla, Heka is more a standalone application rather than a library, but when it comes to acquiring, analyzing, and distributing data such as server logfiles across multiple servers, Heka can prove itself worthy.
Heka is also written in Go, so make sure to check out the source to see how Mozilla utilizes concurrency and Go in real-time data analysis.
Tip
You can visit the Heka home page at http://heka-docs.readthedocs.org/en/latest/ and the Heka source page at https://github.com/mozilla-services/heka.
Finally, there's GoFlow, a flow-based programming paradigm tool that lets you segment your application into distinct components, each capable of being bound to ports, channels, the network, or processes.
While not itself a performance tool, GoFlow might be an appropriate approach to extending concurrency for some applications.
Tip
Visit GoFlow at https://github.com/trustmaster/goflow.