doc: profiling - add README

PR-URL: #339
Reviewed-By: Gireesh Punathil <[email protected]>
Reviewed-By: Michael Dawson <[email protected]>

Author: Peter Marton

documentation/profiling/README.md

@@ -0,0 +1,30 @@
+# Profiling
+
+In this document you can learn about how to profile a Node.js process.
+
+- [Profiling](#profiling)
+  - [My application has a poor performance](#my-application-has-a-poor-performance)
+    - [Symptoms](#symptoms)
+    - [Debugging](#debugging)
+
+## My application has a poor performance
+
+### Symptoms
+
+My applications latency is high and I have already confirmed that the bottleneck
+is not my dependencies like databases and downstream services. So I suspect that
+my application spends significant time to run code or process information.
+
+You are satisfied with your application performance in general but would like to
+understand which part of our application can be improved to run faster or more
+efficient. It can be useful when we want to improve the user experience or save
+computation cost.
+
+### Debugging
+
+In this use-case, we are interested in code pieces that use more CPU cycles than
+the others. When we do this locally, we usually try to optimize our code.
+
+- [Using V8 Profiler](./step1/using_v8_profiler.md)
+- [Using Linux Perf](./step2/using_linux_perf.md.md)
+- [Using Native Tools](./step3/using_native_tools.md)

documentation/profiling/step1/using_v8_profiler.md

@@ -1 +1,9 @@
+# Using V8 Sampling Profiler (--prof)
+
+V8 Profile​ (--prof) is the V8’s built-in sampling based profiler to track CPU
+Usage. ​0x​ is an npm library that makes capturing and visualizing flamegraphs
+via --prof easier.
+
+## How To
+
 //TODO

documentation/profiling/step2/using_linux_perf.md

@@ -0,0 +1,13 @@
+# Using Linux Perf
+
+Linux Perf provides low level CPU profiling with JavaScript, native and OS level frames.
+
+## How To
+
+// TODO
+
+## Useful Links
+
+- https://nodejs.org/en/docs/guides/diagnostics-flamegraph/
+- [http://www.brendangregg.com/blog/2014-09-17/node-flame-graphs-on-linux.html](http://www.brendangregg.com/blog/2014-09-17/node-flame-graphs-on-linux.html)
+- https://github.com/mmarchini/node-linux-perf

documentation/profiling/step2/using_linux_profiler.md

@@ -1 +1,9 @@
+# Native Tools
+
+System profilers (such as Linux perf, eBPF, DTrace, etc.) can sample the stack
+trace of applications and generate reports or graphs to help understand the code
+path our application is taking.
+
+## How To
+
 //TODO