Python Pulumi: Difference between revisions
Line 23: | Line 23: | ||
│ | │ | ||
├─ README.md | ├─ README.md | ||
├─ [[ | ├─ [[Pulumi.yaml#Overview|Pulumi.yaml]] | ||
├─ [[Pulumi_Concepts#Stack_Settings_File|Pulumi.my-stack.yaml]] | ├─ [[Pulumi_Concepts#Stack_Settings_File|Pulumi.my-stack.yaml]] | ||
├─ [[#main_.py|__main__.py]] | ├─ [[#main_.py|__main__.py]] |
Revision as of 23:39, 19 January 2022
External
Internal
Pulumi Python SDK
The SDK is available as a pip package.
Python Pulumi and Virtual Environments
A new Python project created with pulumi new
will have a virtual environment created in a venv
directory with required dependencies from requirements.txt
installed in it. This behavior is controlled by the virtualenv
runtime
option in Pulumi.yaml
.
Manually Initialize Pulumi Python Virtual Environment
Pulumi Python Packges
Adding a New Dependency
To install a new dependency in the virtual environment, add the entry to requirements.txt
and run in the project directory:
venv/bin/pip install -r requirements.txt
Project Layout
The project directory must contain either a __main__.py
file or a setup.py
file that defines the entry point.
my-project │ ├─ README.md ├─ Pulumi.yaml ├─ Pulumi.my-stack.yaml ├─ __main__.py ├─ requirements.txt └─ venv ├─ bin │ ├─ pip │ └─ python ├─ include ├─ lib └─ pyvenv.cfg
Also see:
__main__.py
__main__.py
is the Pulumi Python program that defines the stack resources.
import pulumi
from pulumi_aws import ssm
parameter = ssm.Parameter("/test/a", type="String", value="1234")
pulumi.export("parameter_name", parameter.id)
Python Pulumi Programming Model
Get the name of the currently deploying project:
project = pulumi.get_project()
Getting whether we're doing a dry run or an actual deployment from the program:
project = pulumi.get_project()
Getting a stack programmatically:
stack = pulumi.get_stack()
Environment Variables
All shell environment variables are passed to the running program and can be accessed using the standard runtime APIs os.environ
. This can also be used for dynamic behavior. Configuration is preferable, however, because it makes the behavior less sensitive to local environment configuration.