Note: this is the v3 branch, which is currently in beta. See the docs for v3. If needed the 2.x branch is here, but is in bugfix only mode.
pyinfra turns Python code into shell commands and runs them on your servers. Execute ad-hoc commands and write declarative operations. Target SSH servers, local machine and Docker containers. Fast and scales from one server to thousands. Think ansible
but Python instead of YAML, and a lot faster.
Documentation ⇒ Getting Started • Examples • Help & Support • Contributing
Chat ⇒
#pyinfra
on Matrix
Why pyinfra? Design features include:
- 🚀 Super fast execution over thousands of hosts with predictable performance.
- 🚨 Instant debugging with realtime stdin/stdout/stderr output (
-vvv
). - 🔄 Idempotent operations that enable diffs and dry runs before making changes.
- 📦 Extendable with the entire Python package ecosystem.
- 💻 Agentless execution against anything with shell access.
- 🔌 Integrated with connectors for Docker, Terraform, Vagrant and more.
Install pyinfra with pip
:
pip install pyinfra
Now you can execute commands on hosts via SSH:
pyinfra my-server.net exec -- echo "hello world"
Or target Docker containers, the local machine, and other connectors:
pyinfra @docker/ubuntu exec -- echo "Hello world"
pyinfra @local exec -- echo "Hello world"
As well as executing commands you can define state using operations:
# Install iftop apt package if not present
pyinfra @docker/ubuntu apt.packages iftop update=true _sudo=true
Which can then be saved as a Python file like deploy.py
:
from pyinfra.operations import apt
apt.packages(
name="Ensure iftop is installed",
packages=['iftop'],
update=True,
_sudo=True,
)
The hosts can also be saved in a file, for example inventory.py
:
targets = ["@docker/ubuntu", "my-test-server.net"]
And executed together:
pyinfra inventory.py deploy.py
Now you know the building blocks of pyinfra! By combining inventory, operations and Python code you can deploy anything.
See the more detailed getting started or using operations guides. See how to use inventory & data, global arguments and the CLI or check out the documented examples.