# Profiling

What is flame graph
http://www.brendangregg.com/FlameGraphs/cpuflamegraphs.html

## Profiling on local

若是你的 binary 有開啟 pprof, 並指定 6900 port

import (
	_ "net/http/pprof" //http profiling
)

go tool pprof http://127.0.0.1:6900/debug/pprof/profile -seconds 30

go tool pprof -http=:8081 /Users/jimmy/pprof/pprof.samples.cpu.001.pb.gz


## Profiling on gcp

Just added that on your main binary, then open profiler on gcp console.

// snippets is an example of starting cloud.google.com/go/profiler.
package main

import (
        "cloud.google.com/go/profiler"
)

func main() {
        // Profiler initialization, best done as early as possible.
        if err := profiler.Start(profiler.Config{
                Service:        "myservice",
                ServiceVersion: "1.0.0",
                // ProjectID must be set if not running on GCP.
                // ProjectID: "my-project",
        }); err != nil {
                // TODO: Handle error.
        }
}

### Reference
1. Profiler https://console.cloud.google.com/profiler
2. Real time profiling on gcp
https://medium.com/google-cloud/continuous-profiling-of-go-programs-96d4416af77b
3. GCP profiling
https://cloud.google.com/profiler/docs/profiling-go?hl=zh-TW#gke