mbq.atomiq 0.0.4

MBQ Atomiq

Installation

$ pip install mbq.atomiq

Getting started

1. Add mbq.atomiq to INSTALLED_APPS in your django application’s settings

2. Add ATOMIQ specific settings to that same settings file. Those are used for metrics.

ATOMIQ = {
    'env': CURRENT_ENV,
    'service': YOUR_SERICE_NAME,
}

3. Set up consumers for each queue type that your app needs. mbq.atomiq provides a handy management command for that:

python -m manage atomic_run_consumer --queue sns

python -m manage atomic_run_consumer --queue sqs

python -m manage atomic_run_consumer --queue celery --celery-app path.to.celery.app

notice that celery consumer requires an additional arg celery-app. This should be a module path to the file that instantializes celery app like so:

celery_app = Celery(NAME)

To make sure we’re not holding on to successfully executed or deleted tasks we also have a clean up management command, that by default will clean up all processed tasks that are older than 30 days. That default can be overriden.

python -m manage atomic_cleanup_old_tasks

or

python -m manage atomic_cleanup_old_tasks --days N

or

python -m manage atomic_cleanup_old_tasks --minutes N

4. Use it!

import mbq.atomiq

mbq.atomiq.sns_publish(topic_arn, message)

mbq.atomiq.sqs_publish(queue_url, message)

mbq.atomiq.celery_publish(celery_task, *task_args, **task_kwargs)

Monitoring

<https://app.datadoghq.com/dash/895710/atomiq>

Testing

Tests are automatically in Travis CI https://travis-ci.org/managedbyq/mbq.atomiq but you can also run tests locally using docker-compose. We now use tox for local testing across multiple python environments. Before this use pyenv to install the following python interpreters: cpython{2.7, 3.5, 3.6} and pypy3

$ docker-compose up py36|py27|py37|pypy3

Shipping a New Release

1. Bump the version in __version__.py

2. Go to Releases in GitHub and “Draft a New Release”

3. After creating a new release, Travis CI will pick up the new release and ship it to PyPi