A highly configurable automatic deployment tool integrated with Github. It has limited features and supposed to be used by startups and small organizations.

To simplify your deployment at the early stage of your business and make you focused on building your products.

IDEA -> POC -> Alpha

Read more about our definition of status at starfruit.io.

• Integrated with Github
• Highly Configurable
• Admin UI
• Docker Supported
• Builtin deployment templates to simplify your deployment
• Notification
• Friendly with your existing tool

1. Install crabapple

   pip install crabapple

   If you cloned the repo manually, you can run pip install -U . to install crabapple.

2. Prepare spec for your project

   crabapple new --name demo

   This command will generate a skeleton spec for your project under current folder, tweak that file to match your requirements.

   $ tree demo
   
     demo
     └── spec
   

3. Start crabapple in your server to monitor events from Github

   crabapple server --host 0.0.0.0 --port 50000 --spec demo/spec

   This command will listen at port 50000 and do the auto deployment in terms of the spec.

   If you want it go to the background, run with --daemon.

4. Add a webhook to your Github Repo

   Go to the settings page of your repo, add a webhook, content-type should be application/json.

   URL similiar to: http://xxx.xxx.xxx.xxx:50000/github/event_handler

5. Test it and have fun

   Push something to your master branch and see what's happening.

crabapple subcommand [args]

Use crabapple help subcommand to show the detailed explaination for specific command.

1. new

   This subcommand is used to generate a skeleton deploymeny project, as the optional, you can choose which template to use. The templates there are designed to follow the best practise of the deployment of that specific application to simplify your life.

   Parameters List:

   --name name      : Project name
   [--template tpl] : Template to use
                      django
                      flask
                      tornado
   
   

   Examples:

   crabapple new --name myproj Use generic template to generate the spec file

   crabapple new --name myproj --template django Use django template to generate the spec file

2. server

   This subcommand will launch a HTTP server to monitor events from Github and take actions in terms of your spec. Parameters List:

   -c, --config : Load options from the configuration file
   -h, --host   : Host to bind
   -p, --port   : Port to listen
   -d, --daemon : Run in the background or not
   -s, --spec   : Which spec to use
   --admin      : Enable the Admin UI
   --logdir     : Directory to put all the logs
   --datadir    : Directory To store all the persistent data
   --store      : Persistent way, default to sqlite
   

   Examples:

   crabapple server --host 0.0.0.0 --port 50000 --daemon --spec myproj/spec

   crabapple server --config some_config_file

3. deploy

   This subcommand will do the real deployment by following your deployment spec. It's also used by the crabapple server to do the automatic deployment.

   Parameters List:

   -s, --spec  : Spec file to use
   

   Examples:

   crabapple deploy --spec myproj/spec

Crabapple heavily used this spec to various kinds of stuff. The syntax of it is Python which means you can write any custom functions to extend what you want. Please checkout examples under folder examples.

Simple Example:

# Syntax version of this spec file
syntax_version = '1.0'

# Name of your project
name = 'demo'

# Repo URL of Github
repo_url = 'http://github.com/demo/demo'

# Watched Branches, pushes to these branches will trigger the deployment
watched_branches = ['master']

# Notifiers: email, irc
# notifiers = ['email']

# Email notifier
# notifier_email = {
#     'sender': {
#         'protocol': 'SMTP',
#         'host': 'smtp.gmail.com',
#         'port': 587,
#         'user': '[email protected]',
#         'password': '******',
#     },
#     'recipients': ['[email protected]']
#}

# Called when we try to run the testing
def _on_test():
    # Non-zero and not None means the testing is failed
    print 'run my tests'

on_test = _on_test

# Timeout for one deployment, 0 means no limitations
test_timeout = 0

# Triggered when try to do the deployment
def _on_deploy():
    # Non-zero and not None means the deployment is failed
    print 'run my deployment'

on_deploy = _on_deploy

# Timeout for one deployment, 0 means no limitations
deploy_timeout = 0

An opensource project released by Starfruit.io. We build reuseable open source project for startups.

Feel free to buy us a cup of coffee to encourge us to deliver better open source projects.