Metadata-Version: 2.4
Name: ecs-deplojo
Version: 0.13.0
Summary: Deployment tool for Amazon ECS
Author-email: "Lab Digital B.V." <opensource@labdigital.nl>
License: MIT
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: 3.14
Requires-Python: >=3.11
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: boto3>=1.36.2
Requires-Dist: click>=8.1.8
Requires-Dist: pyaml>=25.1.0
Dynamic: license-file

# ecs-deplojo

[![PyPI](https://img.shields.io/pypi/v/ecs-deplojo)](https://pypi.org/project/ecs-deplojo/)
[![Python Tests](https://github.com/labd/ecs-deplojo/actions/workflows/python-test.yml/badge.svg)](https://github.com/labd/ecs-deplojo/actions/workflows/python-test.yml)

Deployment tool for Amazon ECS.

## Installation

```console
pip install ecs-deplojo
```

## Usage

```console
Usage: ecs-deplojo [OPTIONS]

Options:
  --config FILENAME   [required]
  --var VAR
  --dry-run
  --output-path PATH
  --role-arn <optional arn>
  --help              Show this message and exit.
```

## Example configuration

```yaml
---
cluster_name: example

environment:
    DATABASE_URL: postgresql://

task_definitions:
  web:
    template: task_definitions/web.json
    overrides:
      uwsgi:
        memory: 512
        portMappings:
          - hostPort: 0
            containerPort: 8080
            protocol: tcp
  manage:
    template: task_definitions/manage.json

services:
  web:
    task_definition: web

scheduled_tasks:
  cron-event:
    task_definition: web

before_deploy:
  - task_definition: manage
    container: uwsgi
    command: manage.py migrate --noinput

after_deploy:
  - task_definition: manage
    container: uwsgi
    command: manage.py clearsessions
```

## Using SSM secrets

**Warning secrets will become visible in the container's environent, use with caution**

When you want to use the AWS SSM secrets in your configuration you can use the `secrets`
section, however this needs some additional configuration within AWS

At first you need an AWS IAM role to use as the ECS execution role, this role needs
access to the secrets in Secrets Manager or Parameter store and will only be used during
the startup of your Docker container.

**Example configuration:**

```yaml
--
cluster_name: example

environment:
  NORMAL_ENV_VAR: value_of_variable

secrets:
  DATABASE_URL: /path/to/secret/DATABASE_URL

task_definitions:
  web:
    execution_role_arn: arn:aws:iam::<account_id>:role/execution_role_name
    template: task_definitions/web.json

services:
  web:
    task_definition: web
```

When the container is started the secrets are available as environment variables and
hidden in the AWS ECS console, this is not recommended in production.

## AWS Default VPC

When running your servers in the AWS default VPC you need `networkMode="awsvpc"` in
your task definition JSON file, this will ensure that no hostnames are set for the
containers, since this isn't supported by AWS.

## AWS Fargate

Unlike EC2 based clusters AWS Fargate needs a `execution_role_arn` to work, this can be
set in your service definition in the YAML file.

## AWS Scheduled tasks

The configuration block 'scheduled_tasks' contains the name over the rule in EventBridge
as key, and a reference to the task_definition whose taskDefinitionArn will be used to
update the task definition on the scheduled target.

## Example log output

```console
Starting deploy on cluster example (1 services)
Registered new task definition web:10
Starting one-off task 'manage.py migrate --noinput' via manage:10 (uwsgi)
Updating service web with task defintion web:10
Waiting for deployments
Waiting for services: web (0/2)
Waiting for services: web (1/2)
Waiting for services: web (2/2)
Deployment finished: web (2/2)
Starting one-off task 'manage.py clearsessions' via manage:10 (uwsgi)
```

## Releasing

Releases are managed with [changie](https://changie.dev):

1. Add a changelog entry to your pull request with `changie new`.
2. After merging to `main`, a release pull request is opened (or updated)
   automatically, with the new version and the updated `CHANGELOG.md`.
3. Merging the release pull request creates the GitHub release, which publishes
   the package to PyPI.
