[Django]-Deploying a local django app using openshift

Edit: Remember this is a platform-dependent answer and since the OpenShift platform serving Django may change, this answer could become invalid. As of Apr 1 2016, this answer remains valid at its whole extent.

Many times this happened to me and, since I had to mount at least 5 applications, I had to create my own lifecycle:

1. Don’t use the Django cartridge, but the python 2.7 cartridge. Using the Django cart. and trying to update the django version brings many headaches, not included if you do it from scratch.

2. Clone your repository via git. You will get yourproject and…

   # git clone yourrepo@rhcloud.com:app.git yourproject <- replace it with your actual openshift repo address
   
   yourproject/
   +---wsgi.py
   +---setup.py
   *---.openshift/ (with its contents - I omit them now)
   

3. Make a virtualenv for your brand-new repository cloned into your local machine. Activate it and install Django via pip and all the dependencies you would need (e.g. a new Pillow package, MySQL database package, …). Create a django project there. Say, yourdjproject. Edit Create, alongside, a wsgi/static directory with an empty, dummy, file (e.g. .gitkeep – the name is just convention: you can use any name you want).

    #assuming you have virtualenv-wrapper installed and set-up
    mkvirtualenv myenvironment
    workon myenvironment
    pip install Django[==x.y[.z]] #select your version; optional.
    #creating the project inside the git repository
    cd path/to/yourproject/
    django-admin.py startproject yourjdproject .
    #creating dummy wsgi/static directory for collectstatic
    mkdir -p wsgi/static
    touch wsgi/static/.gitkeep
   

4. Create a django app there. Say, yourapp. Include it in your project.

5. You will have something like this (django 1.7):

   yourproject/
   +---wsgi/
   |   +---static/
   |       +---.gitkeep
   +---wsgi.py
   +---setup.py
   +---.openshift/ (with its contents - I omit them now)
   +---yourdjproject/
   |   +----__init__.py
   |   +----urls.py
   |   +----settings.py
   |   +----wsgi.py
   +---+yourapp/
       +----__init__.py
       +----models.py
       +----views.py
       +----tests.py
       +----migrations
            +---__init__.py
   

6. Set up your django application as you’d always do (I will not detail it here). Remember to include all the dependencies you installed, in the setup.py file accordingly (This answer is not the place to describe WHY, but the setup.py is the package installer and openshift uses it to reinstall your app on each deploy, so keep it up to date with the dependencies).

7. Create your migrations for your models.

8. Edit the openshift-given WSGI script as follows. You will be including the django WSGI application AFTER including the virtualenv (openshift creates one for python cartridges), so the pythonpath will be properly set up.

   #!/usr/bin/python
   import os
   virtenv = os.environ['OPENSHIFT_PYTHON_DIR'] + '/virtenv/'
   virtualenv = os.path.join(virtenv, 'bin/activate_this.py')
   try:
       execfile(virtualenv, dict(__file__=virtualenv))
   except IOError:
       pass
   
   from yourdjproject.wsgi import application
   

9. Edit the hooks in .openshift/action_hooks to automatically perform db sincronization and media management:

   build hook

   #!/bin/bash
   #this is .openshift/action/hooks/build
   #remember to make it +x so openshift can run it.
   if [ ! -d ${OPENSHIFT_DATA_DIR}media ]; then
       mkdir -p ${OPENSHIFT_DATA_DIR}media
   fi
   ln -snf ${OPENSHIFT_DATA_DIR}media $OPENSHIFT_REPO_DIR/wsgi/static/media
   
   ######################### end of file
   

   deploy hook

   #!/bin/bash
   #this one is the deploy hook .openshift/action_hooks/deploy
   source $OPENSHIFT_HOMEDIR/python/virtenv/bin/activate
   cd $OPENSHIFT_REPO_DIR
   echo "Executing 'python manage.py migrate'"
   python manage.py migrate
   echo "Executing 'python manage.py collectstatic --noinput'"
   python manage.py collectstatic --noinput
   
   ########################### end of file
   

10. Now you have the wsgi ready, pointing to the django wsgi by import, and you have your scripts running. It is time to consider the locations for static and media files we used in such scripts. Edit your Django settings to tell where did you want such files:

    STATIC_URL = '/static/'
    MEDIA_URL = '/media/'
    STATIC_ROOT = os.path.join(BASE_DIR, 'wsgi', 'static')
    MEDIA_ROOT = os.path.join(BASE_DIR, 'wsgi', 'static', 'media')
    STATICFILES_DIRS = (os.path.join(BASE_DIR, 'yourjdproject', 'static'),)
    TEMPLATE_DIRS = (os.path.join(BASE_DIR, 'yourjdproject', 'templates'),)
    

11. Create a sample view, a sample model, a sample migration, and PUSH everything.

12. Edit Remember to put the right settings to consider both environments so you can test and run in a local environment AND in openshift (usually, this would involve having a local_settings.py, optionally imported if the file exists, but I will omit that part and put everything in the same file). Please read this file conciously since things like yourlocaldbname are values you MUST set accordingly:

    """
    Django settings for yourdjproject project.
    
    For more information on this file, see
    https://docs.djangoproject.com/en/1.7/topics/settings/
    
    For the full list of settings and their values, see
    https://docs.djangoproject.com/en/1.7/ref/settings/
    """
    
    # Build paths inside the project like this: os.path.join(BASE_DIR, ...)
    import os
    BASE_DIR = os.path.dirname(os.path.dirname(__file__))
    
    ON_OPENSHIFT = False
    if 'OPENSHIFT_REPO_DIR' in os.environ:
        ON_OPENSHIFT = True
    
    
    # Quick-start development settings - unsuitable for production
    # See https://docs.djangoproject.com/en/1.7/howto/deployment/checklist/
    
    # SECURITY WARNING: keep the secret key used in production secret!
    SECRET_KEY = '60e32dn-za#y=x!551tditnset(o9b@2bkh1)b$hn&0$ec5-j7'
    
    # Application definition
    
    INSTALLED_APPS = (
        'django.contrib.admin',
        'django.contrib.auth',
        'django.contrib.contenttypes',
        'django.contrib.sessions',
        'django.contrib.messages',
        'django.contrib.staticfiles',
        'yourapp',
        #more apps here
    )
    
    MIDDLEWARE_CLASSES = (
        'django.contrib.sessions.middleware.SessionMiddleware',
        'django.middleware.common.CommonMiddleware',
        'django.middleware.csrf.CsrfViewMiddleware',
        'django.contrib.auth.middleware.AuthenticationMiddleware',
        'django.contrib.auth.middleware.SessionAuthenticationMiddleware',
        'django.contrib.messages.middleware.MessageMiddleware',
    )
    
    ROOT_URLCONF = 'yourdjproject.urls'
    
    WSGI_APPLICATION = 'yourdjproject.wsgi.application'
    
    # Database
    # https://docs.djangoproject.com/en/1.7/ref/settings/#databases
    
    if ON_OPENSHIFT:
        DEBUG = True
        TEMPLATE_DEBUG = False
        ALLOWED_HOSTS = ['*']
        DATABASES = {
            'default': {
                'ENGINE': 'django.db.backends.mysql',
                'NAME': 'youropenshiftgenerateddatabasename',
                'USER': os.getenv('OPENSHIFT_MYSQL_DB_USERNAME'),
                'PASSWORD': os.getenv('OPENSHIFT_MYSQL_DB_PASSWORD'),
                'HOST': os.getenv('OPENSHIFT_MYSQL_DB_HOST'),
                'PORT': os.getenv('OPENSHIFT_MYSQL_DB_PORT'),
                }
        }
    else:
        DEBUG = True
        TEMPLATE_DEBUG = True
        ALLOWED_HOSTS = []
        DATABASES = {
            'default': {
                'ENGINE': 'django.db.backends.mysql', #If you want to use MySQL
                'NAME': 'yourlocaldbname',
                'USER': 'yourlocalusername',
                'PASSWORD': 'yourlocaluserpassword',
                'HOST': 'yourlocaldbhost',
                'PORT': '3306', #this will be the case for MySQL
            }
        }
    
    # Internationalization
    # https://docs.djangoproject.com/en/1.7/topics/i18n/
    
    LANGUAGE_CODE = 'yr-LC'
    TIME_ZONE = 'Your/Timezone/Here'
    USE_I18N = True
    USE_L10N = True
    USE_TZ = True
    
    # Static files (CSS, JavaScript, Images)
    # https://docs.djangoproject.com/en/1.7/howto/static-files/
    
    STATIC_URL = '/static/'
    MEDIA_URL = '/media/'
    STATIC_ROOT = os.path.join(BASE_DIR, 'wsgi', 'static')
    MEDIA_ROOT = os.path.join(BASE_DIR, 'wsgi', 'static', 'media')
    STATICFILES_DIRS = (os.path.join(BASE_DIR, 'yourdjproject', 'static'),)
    TEMPLATE_DIRS = (os.path.join(BASE_DIR, 'yourdjproject', 'templates'),)
    

13. Git add, commit, push, enjoy.

    cd path/to/yourproject/
    git add .
    git commit -m "Your Message"
    git push origin master # THIS COMMAND WILL TAKE LONG
    # git enjoy
    

14. Your sample Django app is almost ready to go! But if your application has external dependencies it will blow with no apparent reason. This is the reason I told you to develop a simple application. Now it is time to make your dependencies work.

    [untested!] You can edit the deploy hook and add a command after the command cd $OPENSHIFT_REPO_DIR, like this: pip install -r requirements.txt, assuming the requirements.txt file exists in your project. pip should exist in your virtualenv, but if it does not, you can see the next solution.

    Alternatively, the setup.py is an already-provided approach on OpenShift. What I did many times is -assuming the requirements.txt file exists- is:

    1. Open that file, read all its lines.
    2. For each line, if it has a #, remove the # and everything after.
    3. strip leading and trailing whitespaces.
    4. Discard empty lines, and have the result (i.e. remaining lines) as an array.
    5. That result must be assigned to the install_requires= keyword argument in the setup call in the setup.py file.

    I’m sorry I did not include this in the tutorial before! But you need to actually install Django in the server. Perhaps an obvious suggestion, and every Python developer could know that beforehand. But seizing this opportunity I remark: Include the appropriate Django dependency in the requirements.txt (or setup.py depending on whetheryou use or not a requirements.txt file), as you include any other dependency.

This should help you to mount a Django application, and took me a lot of time to standarize the process. Enjoy it and don’t hesitate on contacting me via comment if something goes wrong

Edit (for those with the same problem who don’t expect to find the answer in this post’s comments): Remember that if you edit the build or deploy hook files under Windows and you push the files, they will fly to the server with 0644 permissions, since Windows does not support this permission scheme Unix has, and has no way to assign permissions since these files do not have any extension. You will notice this because your scripts will not be executed when deploying. So try to deploy those files only from Unix-based systems.

Edit 2: You can use git hooks (e.g. pre_commit) to set permissions for certain files, like pipeline scripts (build, deploy, …). See the comments by @StijndeWitt and @OliverBurdekin in this answer, and also this question for more details.