Go程序运行时分析数据：gometry
jopen
10年前
一个可视化界面对来自Go程序运行时数据进行分析。
Installation
go get github.com/rakyll/gometry/cmd/gometryThe program you're willing to profile should import the github.com/rakyll/gometry/http package. The http package will register several handlers to provide information about your program during runtime.
import _ "github.com/rakyll/gometry/http" // If your application is not already running an http server, you need to start one. go func() { log.Println(http.ListenAndServe("localhost:6060", nil)) }() Now, you are ready to launch the gometry.
$ gometryPoint your browser to http://localhost:6464.
Goals
- Building a lightweight tool that works well with live profiles is a necessity. Over the time, I recognized that a lot of people around me delayed to use the existing pprof tools because it's a tedious experience.
- gometry has no ambition to provide the features at the granularity of the features of the command line tools. Users should feel free to fallback to
go tool pprofif they need more sophisticated features. gometry should also provide interfaces to let the users to export their profile data and continue to work with the go tool. - Allow users to filter, hide and ignore by symbol names. Make it easier to generate graphs, make it easier to browse extremely large image files.
- Allow users to work with their custom user profiles.
- Increase the awareness around profiling tools/packages in Go.
- Provide additional lightweight stats where possible.