The most important thing in your depl application is that you create a depl.yml file that contains at least your deploy settings, e.g. deploying Django could look like
deploy:
- django:
settings: settings.prod
On the command line you can use depl deploy myserver.com to deploy your project. You can also add your myserver.com to your settings.
There is extensive documentation for all supported Web Frameworks and Services.
Generally the command line interface looks like this:
Usage:
depl (deploy|remove) [-c=<file>] [-p=<file>] [<host>...]
depl run [-c=<file>] [-p=<file>] <command> [<host>...]
depl -h | --help
Options:
-c, --config=<file> Deploy configuration file [default: depl.yml]
-p, --pool=<name> Define a pool that is going to be deployed.
Only the deploy and run commands are working. The remove command is already reserverd.
While it’s possible to easily define hosts (so you don’t always have to use them while using depl deploy myhost.com):
deploy:
- redis
hosts:
- myhost.com:
password: "the answer is 42"
- host_with_key.com # it is always prefered to use keys.
It’s also possible to group hosts with pools:
deploy:
- postgresql
- django:
port: 8080
hosts:
- databases.com
password: "pg rocks"
- i_dont_like_long_urls.myhost.webservers.com
id: webservers
pools:
db:
deploy: [postgresql]
hosts: [databases.com]
web:
deploy: [django]
hosts: [webservers]
I’m not sure if anybody is ever going to use that, though.
Just use the extends identifier:
deploy:
- redis
extends:
- foo.yml
- bar.yml
Files are automatically merged - it works like multiple inheritance.
There’s two ways of writing custom scripts within depl. The first one is sh , a simple shell script (which shell depends currently on fabric):
deploy:
- sh: |
echo "Hello World"
The second one would be fab. Fab basically resembles the fab files of fabric. It’s writing Python with an from fabric.api import * in front. This is also the way how depl scripts work internally:
deploy:
- fab: |
with warn_only():
sudo('echo "Hello World"')
