--> -->

Platform.sh supports deploying Python applications. Your application can use WSGI-based (Gunicorn / uWSGI) application server, Tornado, Twisted, or Python 3.5+ asyncio server.

Supported 

Grid Dedicated
  • 2.7
  • 3.5
  • 3.6
  • 3.7
  • 3.8
None available

Support libraries 

While it is possible to read the environment directly from your application, it is generally easier and more robust to use the platformshconfig pip library which handles decoding of service credential information for you.

WSGI-based configuration 

In this example, we use Gunicorn to run our WSGI application. Configure the .platform.app.yaml file with a few key settings as listed below, a complete example is included at the end of this section.

  1. Specify the language of your application (available versions are listed above):

    type: 'python:3.8'
  2. Build your application with the build hook. Assuming you have your pip dependencies stored in requirements.txt and a setup.py at the root of your application folder to execute build steps:

    hooks:
      build: |
        pip install -r requirements.txt
        pip install -e .
        pip install gunicorn
    

    These are installed as global dependencies in your environment.

  3. Configure the command you use to start serving your application (this must be a foreground-running process) under the web section, e.g.:

    web:
      commands:
        start: "gunicorn -b 0.0.0.0:$PORT project.wsgi:application"
    

    This assumes the WSGI file is project/wsgi.py and the WSGI application object is named application in the WSGI file.

  4. Define the web locations your application is using:

    web:
      locations:
        "/":
          root: ""
          passthru: true
          allow: false
        "/static":
          root: "static/"
          allow: true
    

    This configuration asks our web server to handle HTTP requests at “/static” to serve static files stored in /app/static/ folder while everything else is forwarded to your application server.

  5. Create any Read/Write mounts. The root file system is read only. You must explicitly describe writable mounts.

    mounts:
        tmp:
            source: local
            source_path: tmp
        logs:
            source: local
            source_path: logs
    

    This setting allows your application writing files to /app/tmp and have logs stored in /app/logs.

Then, set up the routes to your application in .platform/routes.yaml.

"https://{default}/":
  type: upstream
  upstream: "app:http"

Here is the complete .platform.app.yaml file:

name: app
type: python:2.7

web:
  commands:
    start: "gunicorn -b $PORT project.wsgi:application"
  locations:
    "/":
      root: ""
      passthru: true
      allow: false
    "/static":
      root: "static/"
      allow: true

hooks:
  build: |
    pip install -r requirements.txt
    pip install -e .
    pip install gunicorn

mounts:
   tmp:
       source: local
       source_path: tmp
   logs:
       source: local
       source_path: logs

disk: 512

Using the asyncio module 

The above Gunicorn based WSGI example can be modified to use the Python 3.5+ asyncio module.

  1. Change the type to python:3.6.

  2. Change the start command to use asyncio.

    web:
      commands:
        start: "gunicorn -b $PORT -k gaiohttp project.wsgi:application"
    
  3. Add aiohttp as pip dependency in your build hook.

    hooks:
      build: |
        pip install -r requirements.txt
        pip install -e .
        pip install gunicorn aiohttp
    

Accessing services 

To access various services with Python, see the following examples. The individual service pages have more information on configuring each service.

import elasticsearch
from platformshconfig import Config

def usage_example():

    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # Get the credentials to connect to the Elasticsearch service.
    credentials = config.credentials('elasticsearch')

    try:
        # The Elasticsearch library lets you connect to multiple hosts.
        # On Platform.sh Standard there is only a single host so just register that.
        hosts = {
            "scheme": credentials['scheme'],
            "host": credentials['host'],
            "port": credentials['port']
        }

        # Create an Elasticsearch client object.
        client = elasticsearch.Elasticsearch([hosts])

        # Index a few documents
        es_index = 'my_index'
        es_type = 'People'

        params = {
            "index": es_index,
            "type": es_type,
            "body": {"name": ''}
        }

        names = ['Ada Lovelace', 'Alonzo Church', 'Barbara Liskov']

        ids = {}

        for name in names:
            params['body']['name'] = name
            ids[name] = client.index(index=params["index"], doc_type=params["type"], body=params['body'])

        # Force just-added items to be indexed.
        client.indices.refresh(index=es_index)

        # Search for documents.
        result = client.search(index=es_index, body={
            'query': {
                'match': {
                    'name': 'Barbara Liskov'
                }
            }
        })

        table = '''<table>
<thead>
<tr><th>ID</th><th>Name</th></tr>
</thead>
<tbody>'''

        if result['hits']['hits']:
            for record in result['hits']['hits']:
                table += '''<tr><td>{0}</td><td>{1}</td><tr>\n'''.format(record['_id'], record['_source']['name'])
            table += '''</tbody>\n</table>\n'''

        # Delete documents.
        params = {
            "index": es_index,
            "type": es_type,
        }

        for name in names:
            client.delete(index=params['index'], doc_type=params['type'], id=ids[name]['_id'])

        return table

    except Exception as e:
        return e
from json import dumps
from json import loads
from kafka import KafkaConsumer, KafkaProducer
from platformshconfig import Config


def usage_example():
    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()
    # Get the credentials to connect to the Kafka service.
    credentials = config.credentials('kafka')
    
    try:
        kafka_server = '{}:{}'.format(credentials['host'], credentials['port'])
        
        # Producer
        producer = KafkaProducer(
            bootstrap_servers=[kafka_server],
            value_serializer=lambda x: dumps(x).encode('utf-8')
        )
        for e in range(10):
            data = {'number' : e}
            producer.send('numtest', value=data)
        
        # Consumer
        consumer = KafkaConsumer(
            bootstrap_servers=[kafka_server],
            auto_offset_reset='earliest'
        )
        
        consumer.subscribe(['numtest'])
        
        output = ''
        # For demonstration purposes so it doesn't block.
        for e in range(10):
            message = next(consumer)
            output += str(loads(message.value.decode('UTF-8'))["number"]) + ', '

        # What a real implementation would do instead.
        # for message in consumer:
        #     output += loads(message.value.decode('UTF-8'))["number"]

        return output
    
    except Exception as e:
        return e

import pymemcache
from platformshconfig import Config


def usage_example():

    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # Get the credentials to connect to the Memcached service.
    credentials = config.credentials('memcached')

    try:
        # Try connecting to Memached server.
        memcached = pymemcache.Client((credentials['host'], credentials['port']))
        memcached.set('Memcached::OPT_BINARY_PROTOCOL', True)

        key = "Deploy_day"
        value = "Friday"

        # Set a value.
        memcached.set(key, value)

        # Read it back.
        test = memcached.get(key)

        return 'Found value <strong>{0}</strong> for key <strong>{1}</strong>.'.format(test.decode("utf-8"), key)

    except Exception as e:
        return e
from pymongo import MongoClient
from platformshconfig import Config


def usage_example():

    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # The 'database' relationship is generally the name of primary SQL database of an application.
    # It could be anything, though, as in the case here here where it's called "mongodb".
    credentials = config.credentials('mongodb')

    try:
        formatted = config.formatted_credentials('mongodb', 'pymongo')

        server = '{0}://{1}:{2}@{3}'.format(
            credentials['scheme'],
            credentials['username'],
            credentials['password'],
            formatted
        )

        client = MongoClient(server)

        collection = client.main.starwars

        post = {
            "name": "Rey",
            "occupation": "Jedi"
        }

        post_id = collection.insert_one(post).inserted_id

        document = collection.find_one(
            {"_id": post_id}
        )

        # Clean up after ourselves.
        collection.drop()

        return 'Found {0} ({1})<br />'.format(document['name'], document['occupation'])

    except Exception as e:
        return e
import pymysql
from platformshconfig import Config


def usage_example():

    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # The 'database' relationship is generally the name of primary SQL database of an application.
    # That's not required, but much of our default automation code assumes it.'
    credentials = config.credentials('database')

    try:
        # Connect to the database using PDO. If using some other abstraction layer you would inject the values
        # from `database` into whatever your abstraction layer asks for.

        conn = pymysql.connect(host=credentials['host'],
                               port=credentials['port'],
                               database=credentials['path'],
                               user=credentials['username'],
                               password=credentials['password'])

        sql = '''
                CREATE TABLE People (
                id SERIAL PRIMARY KEY,
                name VARCHAR(30) NOT NULL,
                city VARCHAR(30) NOT NULL
                )
                '''

        cur = conn.cursor()
        cur.execute(sql)

        sql = '''
                INSERT INTO People (name, city) VALUES
                ('Neil Armstrong', 'Moon'),
                ('Buzz Aldrin', 'Glen Ridge'),
                ('Sally Ride', 'La Jolla');
                '''

        cur.execute(sql)

        # Show table.
        sql = '''SELECT * FROM People'''
        cur.execute(sql)
        result = cur.fetchall()

        table = '''<table>
<thead>
<tr><th>Name</th><th>City</th></tr>
</thead>
<tbody>'''

        if result:
            for record in result:
                table += '''<tr><td>{0}</td><td>{1}</td><tr>\n'''.format(record[1], record[2])
            table += '''</tbody>\n</table>\n'''

        # Drop table
        sql = '''DROP TABLE People'''
        cur.execute(sql)

        # Close communication with the database
        cur.close()
        conn.close()

        return table

    except Exception as e:
        return e
import psycopg2
from platformshconfig import Config


def usage_example():
    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # The 'database' relationship is generally the name of primary SQL database of an application.
    # That's not required, but much of our default automation code assumes it.' \
    database = config.credentials('postgresql')

    try:
        # Connect to the database.
        conn_params = {
            'host': database['host'],
            'port': database['port'],
            'dbname': database['path'],
            'user': database['username'],
            'password': database['password']
        }

        conn = psycopg2.connect(**conn_params)

        # Open a cursor to perform database operations.
        cur = conn.cursor()

        cur.execute("DROP TABLE IF EXISTS People")

        # Creating a table.
        sql = '''
                CREATE TABLE IF NOT EXISTS People (
                id SERIAL PRIMARY KEY,
                name VARCHAR(30) NOT NULL,
                city VARCHAR(30) NOT NULL
                )
                '''

        cur.execute(sql)

        # Insert data.
        sql = '''
                INSERT INTO People (name, city) VALUES
                ('Neil Armstrong', 'Moon'),
                ('Buzz Aldrin', 'Glen Ridge'),
                ('Sally Ride', 'La Jolla');
                '''

        cur.execute(sql)

        # Show table.
        sql = '''SELECT * FROM People'''
        cur.execute(sql)
        result = cur.fetchall()

        table = '''<table>
<thead>
<tr><th>Name</th><th>City</th></tr>
</thead>
<tbody>'''

        if result:
            for record in result:
                table += '''<tr><td>{0}</td><td>{1}</td><tr>\n'''.format(record[1], record[2])
            table += '''</tbody>\n</table>\n'''

        # Drop table
        sql = "DROP TABLE People"
        cur.execute(sql)

        # Close communication with the database
        cur.close()
        conn.close()

        return table

    except Exception as e:
        return e

import pika
from platformshconfig import Config


def usage_example():
    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # Get the credentials to connect to the RabbitMQ service.
    credentials = config.credentials('rabbitmq')

    try:
        # Connect to the RabbitMQ server
        creds = pika.PlainCredentials(credentials['username'], credentials['password'])
        parameters = pika.ConnectionParameters(credentials['host'], credentials['port'], credentials=creds)

        connection = pika.BlockingConnection(parameters)
        channel = connection.channel()

        # Check to make sure that the recipient queue exists
        channel.queue_declare(queue='deploy_days')

        # Try sending a message over the channel
        channel.basic_publish(exchange='',
                              routing_key='deploy_days',
                              body='Friday!')

        # Receive the message
        def callback(ch, method, properties, body):
            print(" [x] Received {}".format(body))

        # Tell RabbitMQ that this particular function should receive messages from our 'hello' queue
        channel.basic_consume('deploy_days',
                              callback,
                              auto_ack=False)

        # This blocks on waiting for an item from the queue, so comment it out in this demo script.
        # print(' [*] Waiting for messages. To exit press CTRL+C')
        # channel.start_consuming()

        connection.close()

        return " [x] Sent 'Friday!'<br/>"

    except Exception as e:
        return e
from redis import Redis
from platformshconfig import Config


def usage_example():

    # Create a new config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    # Get the credentials to connect to the Redis service.
    credentials = config.credentials('redis')

    try:
        redis = Redis(credentials['host'], credentials['port'])

        key = "Deploy day"
        value = "Friday"

        # Set a value
        redis.set(key, value)

        # Read it back
        test = redis.get(key)

        return 'Found value <strong>{0}</strong> for key <strong>{1}</strong>.'.format(test.decode("utf-8"), key)

    except Exception as e:
        return e

import pysolr
from xml.etree import ElementTree as et
import json
from platformshconfig import Config


def usage_example():

    # Create a new Config object to ease reading the Platform.sh environment variables.
    # You can alternatively use os.environ yourself.
    config = Config()

    try:
        # Get the pysolr-formatted connection string.
        formatted_url = config.formatted_credentials('solr', 'pysolr')

        # Create a new Solr Client using config variables
        client = pysolr.Solr(formatted_url)

        # Add a document
        message = ''
        doc_1 = {
            "id": 123,
            "name": "Valentina Tereshkova"
        }

        result0 = client.add([doc_1], commit=True)
        client.commit()
        message += 'Adding one document. Status (0 is success): {} <br />'.format(json.loads(result0)['responseHeader']['status'])

        # Select one document
        query = client.search('*:*')
        message += '\nSelecting documents (1 expected): {} <br />'.format(str(query.hits))

        # Delete one document
        result1 = client.delete(doc_1['id'])
        client.commit()
        message += '\nDeleting one document. Status (0 is success): {}'.format(et.fromstring(result1)[0][0].text)

        return message

    except Exception as e:
        return e

Project templates 

A number of project templates for Python applications are available on GitHub. Not all of them are proactively maintained but all can be used as a starting point or reference for building your own website or web application.

Basic Python 2

Basic Python 2

This template provides the most basic configuration for running a custom Python 2.7 project. It can be used to build a very rudimentary application but is intended primarily as a documentation reference. The application starts as a bare Python process with no separate runner.

As Python 2.x is deprecated this template should be used for reference only for how to support existing legacy applications. New applications should use a Python 3-based template instead.

Services:

  • Python 2.7.
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Basic Python 3

Basic Python 3

This template provides the most basic configuration for running a custom Python 3.7 project. It includes the `platformshconfig` package and demonstrates using it to connect to MariaDB and Redis. It can be used to build a very rudimentary application but is intended primarily as a documentation reference. The application starts as a bare Python process with no separate runner.

Python is a general purpose scripting language often used in web development.

Services:

  • Python 3.8
  • MariaDB 10.4
  • Redis 5.0
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Django 1

Django 1

This template deploys the Django 1 application framework on Platform.sh, using the gunicorn application runner. It also includes a PostgreSQL database connection pre-configured.

New projects should be built using Django 3, but this project is a reference for existing migrating sites. Version 1 is the legacy support version.

Services:

  • Python 2.7
  • PostgreSQL 10
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Django 2

Django 2

This template deploys the Django 2 application framework on Platform.sh, using the gunicorn application runner. It also includes a PostgreSQL database connection pre-configured.

New projects should be built using Django 3, but this project is a reference for existing migrating sites. Version 2 is in legacy support.

Services:

  • Python 3.8
  • PostgreSQL 12
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Django 3

Django 3

This template deploys the Django 3 application framework on Platform.sh, using the gunicorn application runner. It also includes a PostgreSQL database connection pre-configured.

Django is a Python-based web application framework with a built-in ORM.

Services:

  • Python 3.8
  • PostgreSQL 12
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Flask

Flask

This template demonstrates building the Flask framework for Platform.sh. It includes a minimalist application skeleton that demonstrates how to connect to a MariaDB server for data storage and Redis for caching. The application starts as a bare Python process with no separate runner. It is intended for you to use as a starting point and modify for your own needs.

Flask is a lightweight web microframework for Python.

Services:

  • Python 3.8
  • MariaDB 10.4
  • Redis 5.0
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

MoinMoin

MoinMoin

This template builds a Moin Moin wiki on Platform.sh. The project doesn't include Moin Moin itself; rather, it includes build and deploy scripts that will download Moin Moin on the fly.

Moin Moin is a Python-based Wiki system that uses flat files on disk for storage.

Services:

  • Python 2.7
  • Automatic TLS certificates
  • Moinmoin downloaded on the fly during build

View the repository on GitHub.

Deploy on Platform.sh

Pelican

Pelican

This template provides a basic Pelican skeleton. Only content files need to be committed, as Pelican itself is downloaded at build time via the Pipfile. All files are generated at build time, so at runtime only static files need to be served.

Pelican is a static site generator written in Python and using Jinja for templating.

Services:

  • Python 3.8
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Pyramid

Pyramid

This template builds Pyramid on Platform.sh. It includes a minimalist application skeleton that demonstrates how to connect to a MariaDB server for data storage and Redis for caching. It is intended for you to use as a starting point and modify for your own needs.

Pyramid is a web framework written in Python.

Services:

  • Python 3.8
  • MariaDB 10.4
  • Redis 5.0
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Python 3 running UWSGI

Python 3 running UWSGI

This template provides the most basic configuration for running a custom Python 3.7 project. It includes the `platformshconfig` package and demonstrates using it to connect to MariaDB and Redis. It can be used to build a very rudimentary application but is intended primarily as a documentation reference. The application runs through the UWSGI runner.

Python is a general purpose scripting language often used in web development.

Services:

  • Python 3.8
  • MariaDB 10.4
  • Redis 5.0
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh

Wagtail

Wagtail

This template builds the Wagtail CMS on Platform.sh, using the gunicorn application runner. It includes a PostgreSQL database that is configured automatically, and a basic demonstration app that shows how to use it. It is intended for you to use as a starting point and modify for your own needs. You will need to run the command line installation process by logging into the project over SSH after the first deploy.

Wagtail is a web CMS built using the Django framework for Python.

Services:

  • Python 3.8
  • PostgreSQL 12
  • Automatic TLS certificates
  • Pipfile-based build

View the repository on GitHub.

Deploy on Platform.sh