HomeContact
Computerz
Tips and tricks with ansible
René Dohmen
March 02, 2018
1 min

Recently I’m playing a lot with Ansible. The goal is to deploy a large Django stack to production with all kinds of high availability stuff. The components in this stack: CentOS, Postgres, Nginx with uWSGI and Django, RabbitMQ, several celery workers with supervisord. As a subgoal it would be nice to run the complete production environment also from inside a vagrant VM; but then with all components on a single host.

Deploying a Django stack on CentOS 6.5

A small, but complete, prototype can be found [here](https://github.com/acidjunk/ansible-centos-django-stack). It’s based on a fork of [this](https://github.com/tcosta84/ansible-centos-django-stack)

You can try this complete Django setup, when you already have Ansible, VirtualBox and Vagrant installed by running:

Working on your Ansible playbooks

I use Sublime for editing. Because i’m used to PyCharms autosave I installed the Sublime’s autosave plugin for that. Sublime has excellent YAML syntax highlighting out of the box.

Some handy commands

As a quick reminder for myself I list some commands here that I used a lot while working on the Ansible scripts.

Run commands on your cluster

When you have an inventory file with the group all:

Show all ansible setup variabels inside your cluster

It can be quite handy to see the ansible variabeles/environment stuff on a server or VM:

Quircks with handlers

Keep in mind that ansible only calls the handlers upon successful completion of the completion of the playbook. When a task fails on a host, handlers which were previously notified will not be run on that host. This can lead to cases where an unrelated failure can leave a host in an unexpected state. For example, a task could update a configuration file and notify a handler to restart some service. If a task later on in the same play fails, the service will not be restarted despite the configuration change. You could circumvent that by providing [--force-handlers]{.title-ref} on the command line. If it happened while working on a playbook you can trigger that tasks inside the handler clause manually by running:

Start clean

When you need to start clean again you can do this:

Other handy adhoc stuff

Install a package on CentOS hosts:

Deinstall it again:

Add a firewalld rule, activate it now and after the next boot make it permanent:


Related Posts

Deploy FastAPI on Amazon Serverless
May 15, 2021
5 min
© 2021, All Rights Reserved.
Powered by formatics

Quick Links

Advertise with usAbout UsContact Us

Social Media