Plugins

Sentry

install

We can use sentry to monitor our application.

To install sentry, we can download zip file on the server, then unzip it.

Then, ensure that docker is installed, go to the repo and run ./install.sh

sentry needs 4G of RAM at least.

After installation, run docker-compose up -d, then we can use sentry on 127.0.0.1:9000

create a sentry project

In settings of sentry app, first, we create a Team, then, Project.

After than, sentry will tell us how to config our app to use sentry.

usage of sentry

Sentry can collect automatically:

  • uncaught error
  • error log (logger.error)

We can also send our custom message / error to sentry using capture_message / capture_exception.

Sentry creates many SDK for different frameworks and languages.

REST framework

official

It opens api to outside. To use it, we need:

  1. install packages

  2. config settings.py to enable REST framework and set the authentication

  3. config urls.py to create:

    • Serializers

    • ViewSets of api pages

    • Routers of api

Redis

Redis is cache tool.

Before starting, we need to install redis on the host.

Then install django-redis. offical

Follow the official docs to set up django-redis, then add middleware:

# setting.py

MIDDLEWARE =[
    ...,
    'django.middleware.cache.UpdateCacheMiddleware',
    'django.middleware.common.CommonMiddleware',
    'django.middleware.cache.FetchFromCacheMiddleware',
    ...
]

The order is important. That means, for each view, django always,

  1. searches it in the cache
  2. do something common
  3. update the cache

Celery

Celery is a simple, flexible, and reliable distributed system. It’s a task queue with focus on real-time processing.

official

Installation

We need to install Celery and its Bundles, such as redis,auth, msgpack

First step

example

In this example, we create a simple task. We set redis as the broker, and we save the task results in redis

  1. run a worker server

    To run it: celery -A tasks worker --loglevel=INFO

    That means we execute our app (-A), named tasks with the worker. tasks is the name of the python file.

  2. call the task

    We create another script to call add function in tasks.py.

  3. monitor the tasks

    We use Flower to monitor.

    After installing, run monitor of the task:

    • celery -A tasks flower

Celery with django

official

There are some custom settings for recruitment project:

# recruitment/recruitment/celery.py

# set the default Django settings module for the 'celery' program.
os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'settings.base')

app = Celery('recruitment')

To make it work:

  1. start redis:

    redis-server

  2. start the worker:

    DJANGO_SETTINGS_MODULE=settings.local celery -A recruitment worker --loglevel=INFO

  3. start the monitor (optional):

    DJANGO_SETTINGS_MODULE=settings.local celery -A recruitment flower

Celery with django beat

official

This extension enables you to store the periodic task schedule in the database.

install

  1. intall pip install django-celery-beat

  2. add django_celery_beat into INSTALLED_APPS in settings

  3. migrate database: python manage.py migrate

  4. start beat

    DJANGO_SETTINGS_MODULE=settings.local celery -A recruitment beat --scheduler django_celery_beat.schedulers:DatabaseScheduler

    It tells beat to store periodic task schedule in the database. (--scheduler django_celery_beat.schedulers:DatabaseScheduler)

add Periodic Tasks

doc

Other plugins

  • Django debug toolbar

  • django-silk: analyze performance bottleneck

  • haystack django: search

  • django notifications

  • django markdown editor

  • simple UI: django Admin UI

  • django-crispy-forms: forms UI

  • django-simple-captcha