Posted on 2012-12-27 13:45:00+00:00
Fabric is an excellent tool you can use to deploy your sites and projects to servers. Though it's written in Python, you can use it to deploy just about anything. Typically, the Fabric execution model revolves around a remote server, and your own local machine. But what if you want to perform an operation between two servers? This post will show you one approach.
First, as usual, our assumptions:
With that out of the way, let's start.
For this tutorial, we're going to define a very basic backpublishing task. That means we're going to copy a database from a production server to a staging server. So create a
fabfile.py and let's define our file structure:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
As you can see, our actual task is split up into two parts. And there lies the key to this technique: you split up the operation into the parts that need to be executed on each server. Let's fill those parts out:
1 2 3 4 5
Remember: we don't combine these two parts into one function because they need to
be executed on different servers. Let's fill out the main task to show you how
to do it:
1 2 3 4 5 6 7
And that's it! Now the different parts will be executed on different servers, without any low level hacks, or having to split your workflow into separate Fabric tasks. Of course, we're pulling in the role functionality into the actual Fabric task, so you'd call this from bash like so:
For reference, here's the entire
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20