So, you have a hundred VMs you need to manage, and you have ... Ansible ? I should stop here, as this is a tool that is standard in configuration management nowadays, but I'd dare to continue and say there is a better alternative to it.

But before we get into it, why am I so frustrated with Ansible? Here are my points:

• YAML based declarative DSL really stinks on complex tasks as it lacks the flexibility that imperative languages have.

• YAML is not even a programming language, and you gonna pay the price very soon.

• To keep ansible code clean and simple, extra efforts are required, one need to refactor out all the complexity from YAML to python modules and this feels like "why I even start using YAML DSL"?

• Ansible reports are frustrating as I always need to add these debug tasks to show real STDOUT/STDERR emitted from commands, where it should just work out of the box.

• Ansible ties me to the idea of "running on a host," where sometimes I need to run tasks not tied to hosts, yes, you can still use "ansible_connection=local" but this feels awkward.

Meet Sparky

So, meet Sparky - elegant, efficient and all-battery included automation tool. It's written on powerful and modern Raku language, with bulma css frontend and web sockets.

To install Sparky - install Rakudo first and then install Sparky itself as a Raku module:

curl https://rakubrew.org/install-on-perl.sh | sh 
eval "$(~/.rakubrew/bin/rakubrew init Bash)" 
rakubrew download moar-2024.05
git clone https://github.com/melezhik/sparky.git
cd sparky/
# install Sparky and it's dependencies
zef install --/test .
# init sparky sqlite database
raku db-init.raku
# run sparky job runner
nohup sparkyd >~/.sparkyd.log < /dev/null &
# run sparky web console
cro run

This simple scenario gets it up and running; if you go to http://127.0.0.1:4000 you'll see a nice Sparky web console. We use the console to run sparky jobs.

Show me the design

So we have a control plane that would manage many hosts over ssh, using push mode:

         ---------------
         | CP , Sparky | 
         ---------------
              [ssh] 
       /   /    |    \    \
     host host host host host

This is pretty much what ansible does ...

Show me the code

Now say, we have 5 NGINX servers we need to restart, let's drop a simple Sparky job to do this in pure Raku language:

use Sparky::JobApi;
class Pipeline does Sparky::JobApi::Role { 
  method stage-main {
    for 1..5 -> $i {
      my $j = self.new-job :workers<5>; 
      $j.queue: %(
        sparrowdo => %(
          bootstrap => true,
          host => "nginx_{$i}.local.domain"
        ),
       tags => %(
         stage => "child",
         i => $i
       )
      ); 
   }
  }
  method stage-child {
    service-restart "nginx" 
  }
}

In this scenario, Sparky will run five parallel jobs that restart nginx on five hosts. Simple and elegant.

Moreover those five jobs will appear as five separate reports in Sparky UI …

Got interested?

Of course, this is only a quick glance at Sparky architecture and features, things to cover further:

• Job orchestration (DAGs)
• Core DSL (pure Raku)
• Custom UIs
• Authentication (oauth2) and security access list
• Writing more sophisticated scenarios
• Extending Sparky with plugins by using many programming languages
• Installing Sparky with MySQL/Postgresql database storage (instead of sqlite)
• Using Sparky as CI server (SCM triggering and cron jobs)

Links

Sparky project - https://github.com/melezhik/sparky