AnsibleMetrics is a Python-based static source code measurement tool to characterize Infrastructure-as-Code. It helps quantify the characteristics of infrastructure code to support DevOps engineers when maintaining and evolving it. It currently supports 46 source code metrics, though other metrics can be derived by combining the implemented ones.
If you use AnsibleMetrics in a scientific publication, we would appreciate citations to the following paper:
@article{DALLAPALMA2020100633,
title = "AnsibleMetrics: A Python library for measuring Infrastructure-as-Code blueprints in Ansible",
journal = "SoftwareX",
volume = "12",
pages = "100633",
year = "2020",
issn = "2352-7110",
doi = "https://doi.org/10.1016/j.softx.2020.100633",
url = "http:https://www.sciencedirect.com/science/article/pii/S2352711020303460",
author = "Stefano {Dalla Palma} and Dario {Di Nucci} and Damian A. Tamburri",
keywords = "Infrastructure as Code, Software metrics, Software quality",
abstract = "Infrastructure-as-Code (IaC) has recently received increasing attention in the research community, mainly due to the paradigm shift it brings in software design, development, and operations management. However, while IaC represents an ever-increasing and widely adopted practice, concerns arise about the need for instruments that help DevOps engineers efficiently maintain, speedily evolve, and continuously improve Infrastructure-as-Code. In this paper, we present AnsibleMetrics, a Python-based static source code measurement tool to characterize Infrastructure-as-Code. Although we focus on Ansible, the most used language for IaC, our tool could be easily extended to support additional formats. AnsibleMetrics represents a step forward towards software quality support for DevOps engineers developing and maintaining infrastructure code."
}
Installation is made simple by the PyPI repository. Download the tool and install it with:
pip install ansiblemetrics
or, alternatively from the source code project directory:
pip install -r requirements.txt
pip install .
Run ansible-metrics --help
for instructions about the usage:
usage: ansible-metrics [-h] [--omit-zero-metrics] [-d DEST] [-o] [-v] src
Extract metrics from Ansible scripts.
positional arguments:
src source file (playbook or tasks file) or
directory
optional arguments:
-h, --help show this help message and exit
--omit-zero-metrics omit metrics with value equal 0
-d DEST, --dest DEST destination path to save results
-o, --output shows output
-v, --version show program's version number and exit
Assume that the following example is named playbook1.yml:
---
- hosts: webservers
vars:
http_port: 80
remote_user: root
tasks:
- name: ensure apache is at the latest version
yum:
name: httpd
state: latest
- hosts: databases
remote_user: root
tasks:
- name: ensure postgresql is at the latest version
yum:
name: postgresql
state: latest
- name: ensure that postgresql is started
service:
name: postgresql
state: started
and is located within the folder playbooks as follows:
playbooks
|- playbook1.yml
|- playbook2.yml
|- playbook3.yml
Also, assume the user's working directory is the playbooks folder. Then, it is possible to extract source code characteristics from that blueprint by running the following command:
ansible-metrics --omit-zero-metrics playbook1.yml --dest report.json
For this example, the report.json will result in
{
"filepath": "playbook1.yml",
"avg_play_size": 10,
"avg_task_size": 4,
"lines_blank": 4,
"lines_code": 20,
"num_keys": 20,
"num_parameters": 6,
"num_plays": 2,
"num_tasks": 3,
"num_tokens": 50,
"num_unique_names": 3,
"num_vars": 1,
"text_entropy": 4.37
}
AnsibleMetrics currently supports up to 46 source code metrics, implemented in Python. To extract the value for a given metric follow this pattern:
from ansiblemetrics.<general|playbook>.metric import Metric
script = 'a valid yaml script'
value = Metric(script).count()
where metric and Metric have to be replaced with the name of the desired metric module to compute the value of a specific metric.
The difference between the general and the playbook modules lies in the fact that the playbook module contains metrics specific to playbooks (for example, the number of plays and tasks), while the general module contains metrics that can be generalized to other languages (for example, the lines of code).
For example, to count the number of lines of code:
from ansiblemetrics.general.lines_code import LinesCode
script = """
---
- hosts: all
tasks:
- name: This is a task!
debug:
msg: "Hello World"
"""
print('Lines of executable code:', LinesCode(script).count())
To extract the value for the 46 metrics at once, import the ansiblemetrics.metrics_extractor
package and call the method extract_all()
(in this case the return value will be a json object):
from ansiblemetrics.metrics_extractor import extract_all
script = """
---
- hosts: all
tasks:
- name: This is a task!
debug:
msg: "Hello World"
"""
metrics = extract_all(script)
print('Lines of executable code:', metrics['lines_code'])