https://github.com/marwan-at-work/minicd

A Go HandlerFunc That Makes Your Go Server Self-Deployable
https://github.com/marwan-at-work/minicd

continuous-delivery golang

Last synced: 2 days ago
JSON representation

A Go HandlerFunc That Makes Your Go Server Self-Deployable

• Host: GitHub
• URL: https://github.com/marwan-at-work/minicd
• Owner: marwan-at-work
• Created: 2017-07-29T23:00:07.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-07-30T15:03:07.000Z (over 7 years ago)
• Last Synced: 2024-11-06T01:57:49.573Z (about 2 months ago)
• Topics: continuous-delivery, golang
• Language: Go
• Homepage:
• Size: 3.66 MB
• Stars: 1
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# MiniCD

Package minicd exposes an `http.HandlerFunc` that you can include in your Go web server so it can continuously deliver itself whenever you push to Github. 

This is meant for small personal projects and I wouldn't recommend it for large production apps. For larger scale CI/CD check out [baghdad](https://www.github.com/marwan-at-work/baghdad).

## Usage

```go

package main

import (

    "context"

    "fmt"

    "net/http"

    "github.com/marwan-at-work/minicd"

)

func main() {

    killSig := make(chan context.Context)

    http.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {

        fmt.Fprintln(w, "hello world")

    })

    http.HandleFunc("/github-hook", minicd.Handler(minicd.Config{

        WebhookSecret: "my_webhook_secret",

        GithubToken: "optional_if_public_repo",

        KillSig: killSig,

    }))

    srv := &http.Server{Addr: ":3000"}

    go func() {

        if err := srv.ListenAndServe(); err != nil && err != http.ErrServerClosed {

            log.Fatal(err)

        }

    }()

    err := srv.Shutdown(<-killSig)

    if err != nil {

        log.Fatal(err)

    }

    fmt.Println("see ya on the other updated side")

}

```

### Gotchas

This doesn't work with a local Docker container because spinning a new process inside the container doesn't keep the container alive if the original process was killed. 

As a to-do, this could work with a swarm cluster.