Welcome to django-tenants documentation!

This application enables Django powered websites to have multiple tenants via PostgreSQL schemas. A vital feature for every Software-as-a-Service website.

Django provides currently no simple way to support multiple tenants using the same project instance, even when only the data is different. Because we don’t want you running many copies of your project, you’ll be able to have:

  • Multiple customers running on the same instance

  • Shared and Tenant-Specific data

  • Tenant View-Routing

What are schemas?

A schema can be seen as a directory in an operating system, each directory (schema) with it’s own set of files (tables and objects). This allows the same table name and objects to be used in different schemas without conflict. For an accurate description on schemas, see PostgreSQL’s official documentation on schemas.

Why schemas?

There are typically three solutions for solving the multitenancy problem.

  1. Isolated Approach: Separate Databases. Each tenant has it’s own database.

  2. Semi Isolated Approach: Shared Database, Separate Schemas. One database for all tenants, but one schema per tenant.

  3. Shared Approach: Shared Database, Shared Schema. All tenants share the same database and schema. There is a main tenant-table, where all other tables have a foreign key pointing to.

This application implements the second approach, which in our opinion, represents the ideal compromise between simplicity and performance.

  • Simplicity: barely make any changes to your current code to support multitenancy. Plus, you only manage one database.

  • Performance: make use of shared connections, buffers and memory.

Each solution has it’s up and down sides, for a more in-depth discussion, see Microsoft’s excellent article on Multi-Tenant Data Architecture.

How it works

Tenants are identified via their host name (i.e tenant.domain.com). This information is stored on a table on the public schema. Whenever a request is made, the host name is used to match a tenant in the database. If there’s a match, the search path is updated to use this tenant’s schema. So from now on all queries will take place at the tenant’s schema. For example, suppose you have a tenant customer at http://customer.example.com. Any request incoming at customer.example.com will automatically use customer’s schema and make the tenant available at the request. If no tenant is found, a 404 error is raised. This also means you should have a tenant for your main domain, typically using the public schema. For more information please read the [setup](#setup) section.

Important

Tenant’s domain name and schema name are usually the same or similar but they don’t have to be! For example the tenant at http://acme.example.com could be backed by the acme schema, while http://looney-tunes.tld could be backed by the tenant2 schema! Notice that domain names are not related in any way to schema names! There is also no restriction whether or not you should use sub-domains or top-level domains.

Warning

Schema names and domain names have different validation rules. Underscores (_) and capital letters are permitted in schema names but they are illegal for domain names! On the other hand domain names may contain a dash (-) which is illegal for schema names!

You must be careful if using schema names and domain names interchangeably in your multi-tenant applications! The tenant and domain model classes, creation and validation of input data are something that you need to handle yourself, possibly imposing additional constraints to the acceptable values!

Shared and Tenant-Specific Applications

Tenant-Specific Applications

Most of your applications are probably tenant-specific, that is, its data is not to be shared with any of the other tenants.

Shared Applications

An application is considered to be shared when its tables are in the public schema. Some apps make sense being shared. Suppose you have some sort of public data set, for example, a table containing census data. You want every tenant to be able to query it. This application enables shared apps by always adding the public schema to the search path, making these apps also always available.

Contents

Installation

Assuming you have django installed, the first step is to install django-tenants.

pip install django-tenants

Basic Settings

You’ll have to make the following modifications to your settings.py file.

Your DATABASE_ENGINE setting needs to be changed to

DATABASES = {
    'default': {
        'ENGINE': 'django_tenants.postgresql_backend',
        # ..
    }
}

Add django_tenants.routers.TenantSyncRouter to your DATABASE_ROUTERS setting, so that the correct apps can be synced, depending on what’s being synced (shared or tenant).

DATABASE_ROUTERS = (
    'django_tenants.routers.TenantSyncRouter',
)

Add the middleware django_tenants.middleware.main.TenantMainMiddleware to the top of MIDDLEWARE, so that each request can be set to use the correct schema.

MIDDLEWARE = (
    'django_tenants.middleware.main.TenantMainMiddleware',
    #...
)

Make sure you have django.template.context_processors.request listed under the context_processors option of TEMPLATES otherwise the tenant will not be available on request.

TEMPLATES = [
    {
        #...
        'OPTIONS': {
            'context_processors': [
                'django.template.context_processors.request',
                #...
            ],
        },
    },
]

The Tenant & Domain Model

Now we have to create your tenant model. Your tenant model can contain whichever fields you want, however, you must inherit from TenantMixin. This Mixin only has one field schema_name which is required. You also have to have a table for your domain names for this you must inherit from DomainMixin .

Here’s an example, suppose we have an app named customers and we want to create a model called Client.

from django.db import models
from django_tenants.models import TenantMixin, DomainMixin

class Client(TenantMixin):
    name = models.CharField(max_length=100)
    paid_until =  models.DateField()
    on_trial = models.BooleanField()
    created_on = models.DateField(auto_now_add=True)

    # default true, schema will be automatically created and synced when it is saved
    auto_create_schema = True

class Domain(DomainMixin):
    pass

Admin Support

TenantAdminMixin is available in order to register the tenant model. Here’s an example (following the example above), we want to register the Client model, so we create a the related admin class ClientAdmin. The mixin disables save and delete buttons when not in current or public tenant (preventing Exceptions).

from django.contrib import admin
from django_tenants.admin import TenantAdminMixin

from myapp.models import Client

@admin.register(Client)
class ClientAdmin(TenantAdminMixin, admin.ModelAdmin):
        list_display = ('name', 'paid_until')

Configure Tenant and Shared Applications

To make use of shared and tenant-specific applications, there are two settings called SHARED_APPS and TENANT_APPS. SHARED_APPS is a tuple of strings just like INSTALLED_APPS and should contain all apps that you want to be synced to public. If SHARED_APPS is set, then these are the only apps that will be synced to your public schema! The same applies for TENANT_APPS, it expects a tuple of strings where each string is an app. If set, only those applications will be synced to all your tenants. Here’s a sample setting

SHARED_APPS = (
    'django_tenants',  # mandatory
    'customers', # you must list the app where your tenant model resides in

    'django.contrib.contenttypes',

    # everything below here is optional
    'django.contrib.auth',
    'django.contrib.sessions',
    'django.contrib.sites',
    'django.contrib.messages',
    'django.contrib.admin',
)

TENANT_APPS = (
    # your tenant-specific apps
    'myapp.hotels',
    'myapp.houses',
)

INSTALLED_APPS = list(SHARED_APPS) + [app for app in TENANT_APPS if app not in SHARED_APPS]

You also have to set where your tenant & domain models are located.

TENANT_MODEL = "customers.Client" # app.Model

TENANT_DOMAIN_MODEL = "customers.Domain"  # app.Model

Now run migrate_schemas --shared, this will create the shared apps on the public schema. Note: your database should be empty if this is the first time you’re running this command.

python manage.py migrate_schemas --shared

Note

If you use migrate migrations will be applied to both shared and tenant schemas!

Warning

You might need to run makemigrations and then migrate_schemas --shared again for your app.Models to be created in the database.

Lastly, you need to create a tenant whose schema is public and it’s address is your domain URL. Please see the section on use.

You can also specify extra schemas that should be visible to all queries using PG_EXTRA_SEARCH_PATHS setting.

PG_EXTRA_SEARCH_PATHS = ['extensions']

PG_EXTRA_SEARCH_PATHS should be a list of schemas you want to make visible globally.

Tip

You can create a dedicated schema to hold postgresql extensions and make it available globally. This helps avoid issues caused by hiding the public schema from queries.

Warning

By default, a check is performed to validate that the additional extension schemas do not conflict with the tenant schemas. If you want to ignore this check, you can set SKIP_PG_EXTRA_VALIDATION to True. We do not recommend disabling this validation, it is at your own risk.

Sub-folder Support

Currently in beta.

There is a option that allows you to run Django-Tenants with sub-folder instead of sub-domains.

Warning

The schemaname value from the URL path has to match domain value which is set when creating a new tenant, as described https://django-tenants.readthedocs.io/en/latest/use.html#creating-a-tenant

TENANT_SUBFOLDER_PREFIX needs to be added to the settings file. This is the url prefix for the tenant this can’t be left blank.

TENANT_SUBFOLDER_PREFIX = "clients"

In the example given above, the prefixed path ``/r`` will become ``/clients``.
e.g.  http://www.mydomain.local/clients/schemaname/ instead of http://www.mydomain.local/r/schemaname/

The middleware is different to the standard middleware. The middleware required is

MIDDLEWARE = (
    'django_tenants.middleware.TenantSubfolderMiddleware',
    #...
)

Tip

There is an example project for this in the examples folder

Optional Settings

PUBLIC_SCHEMA_NAME
Default

'public'

The schema name that will be treated as public, that is, where the SHARED_APPS will be created.

TENANT_CREATION_FAKES_MIGRATIONS
Default

'False'

Sets if the schemas will be copied from an existing “template” schema instead of running migrations. Useful in the cases where migrations can not be faked and need to be ran individually, or when running migrations takes a long time. Be aware that setting this to True may significantly slow down the process of creating tenants.

When using this option, you must also specify which schema to use as template, under TENANT_BASE_SCHEMA.

TENANT_BASE_SCHEMA
Default

None

The name of the schema to use as a template for creating new tenants. Only used when TENANT_CREATION_FAKES_MIGRATIONS is enabled.

TENANT_SYNC_ROUTER
Default

django_tenants.routers.TenantSyncRouter

The name of the database router that ready() checks for when the django_tenant app checks for. If set then place this in DATABASE_ROUTERS.

DATABASE_ROUTERS = [
    # ..
    TENANT_SYNC_ROUTER
    # ..
]
TENANT_MIGRATION_ORDER
Default

None

A list of fields to order the tenant queryset by when migrating schemas.

Tenant View-Routing
PUBLIC_SCHEMA_URLCONF
Default

None

We have a goodie called PUBLIC_SCHEMA_URLCONF. Suppose you have your main website at example.com and a customer at customer.example.com. You probably want your user to be routed to different views when someone requests http://example.com/ and http://customer.example.com/. Because django only uses the string after the host name, this would be impossible, both would call the view at /. This is where PUBLIC_SCHEMA_URLCONF comes in handy. If set, when the public schema is being requested, the value of this variable will be used instead of ROOT_URLCONF. So for example, if you have

PUBLIC_SCHEMA_URLCONF = 'myproject.urls_public'

When requesting the view /login/ from the public tenant (your main website), it will search for this path on PUBLIC_SCHEMA_URLCONF instead of ROOT_URLCONF.

Separate projects for the main website and tenants (optional)

In some cases using the PUBLIC_SCHEMA_URLCONF can be difficult. For example, Django CMS takes some control over the default Django URL routing by using middlewares that do not play well with the tenants. Another example would be when some apps on the main website need different settings than the tenants website. In these cases it is much simpler if you just run the main website example.com as a separate application.

If your projects are ran using a WSGI configuration, this can be done by creating a file called wsgi_main_website.py in the same folder as wsgi.py.

# wsgi_main_website.py
import os
os.environ.setdefault("DJANGO_SETTINGS_MODULE", "project.settings_public")

from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()

If you put this in the same Django project, you can make a new settings_public.py which points to a different urls_public.py. This has the advantage that you can use the same apps that you use for your tenant websites.

Or you can create a completely separate project for the main website.

Caching

To enable tenant aware caching you can set the KEY_FUNCTION setting to use the provided make_key helper function which adds the tenants schema_name as the first key prefix.

CACHES = {
    "default": {
        ...
        'KEY_FUNCTION': 'django_tenants.cache.make_key',
        'REVERSE_KEY_FUNCTION': 'django_tenants.cache.reverse_key',
    },
}

The REVERSE_KEY_FUNCTION setting is only required if you are using the django-redis cache backend.

Configuring your Apache Server (optional)

Here’s how you can configure your Apache server to route all subdomains to your django project so you don’t have to setup any subdomains manually.

<VirtualHost 127.0.0.1:8080>
    ServerName mywebsite.com
    ServerAlias *.mywebsite.com mywebsite.com
    WSGIScriptAlias / "/path/to/django/scripts/mywebsite.wsgi"
</VirtualHost>

Django’s Deployment with Apache and mod_wsgi might interest you too.

Building Documentation

Documentation is available in docs and can be built into a number of formats using Sphinx. To get started

pip install Sphinx
cd docs
make html

This creates the documentation in HTML format at docs/_build/html.

Using django-tenants

Creating a Tenant

Creating a tenant works just like any other model in django. The first thing we should do is to create the public tenant to make our main website available. We’ll use the previous model we defined for Client.

from customers.models import Client, Domain

# create your public tenant
tenant = Client(schema_name='public',
                name='Schemas Inc.',
                paid_until='2016-12-05',
                on_trial=False)
tenant.save()

# Add one or more domains for the tenant
domain = Domain()
domain.domain = 'my-domain.com' # don't add your port or www here! on a local server you'll want to use localhost here
domain.tenant = tenant
domain.is_primary = True
domain.save()

Now we can create our first real tenant.

from customers.models import Client, Domain

# create your first real tenant
tenant = Client(schema_name='tenant1',
                name='Fonzy Tenant',
                paid_until='2014-12-05',
                on_trial=True)
tenant.save() # migrate_schemas automatically called, your tenant is ready to be used!

# Add one or more domains for the tenant
domain = Domain()
domain.domain = 'tenant.my-domain.com' # don't add your port or www here!
domain.tenant = tenant
domain.is_primary = True
domain.save()

Because you have the tenant middleware installed, any request made to tenant.my-domain.com will now automatically set your PostgreSQL’s search_path to tenant1, public, making shared apps available too. The tenant will be made available at request.tenant. By the way, the current schema is also available at connection.schema_name, which is useful, for example, if you want to hook to any of django’s signals.

Any call to the methods filter, get, save, delete or any other function involving a database connection will now be done at the tenant’s schema, so you shouldn’t need to change anything at your views.

Deleting a tenant

You can delete tenants by just deleting the entry via the Django ORM. There is a flag that can set on the tenant model called auto_drop_schema. The default for auto_drop_schema is False. WARNING SETTING AUTO_DROP_SCHEMA TO TRUE WITH DELETE WITH TENANT!

Utils

There are several utils available in django_tenants.utils that can help you in writing more complicated applications.

schema_context(schema_name)

This is a context manager. Database queries performed inside it will be executed in against the passed schema_name. (with statement)

from django_tenants.utils import schema_context

with schema_context(schema_name):
    # All commands here are ran under the schema `schema_name`

# Restores the `SEARCH_PATH` to its original value

You can also use schema_context as a decorator.

from django_tenants.utils import schema_context

@schema_context(schema_name)
def my_func():
  # All commands in this function are ran under the schema `schema_name`
tenant_context(tenant_object)

This context manager is very similar to the schema_context function, but it takes a tenant model object as the argument instead.

from django_tenants.utils import tenant_context

with tenant_context(tenant):
    # All commands here are ran under the schema from the `tenant` object

# Restores the `SEARCH_PATH` to its original value

You can also use tenant_context as a decorator.

from django_tenants.utils import tenant_context

@tenant_context(tenant)
def my_func():
  # All commands in this function are ran under the schema from the `tenant` object

Signals

There are number of signals

`post_schema_sync` will get called after a schema gets created from the save method on the tenant class.

`schema_needs_to_be_sync` will get called if the schema needs to be migrated. `auto_create_schema` (on the tenant model) has to be set to False for this signal to get called. This signal is very useful when tenants are created via a background process such as celery.

`schema_migrated` will get called once migrations finish running for a schema.

`schema_migrate_message` will get called after each migration with the message of the migration. This signal is very useful when for process / status bars.

Example

@receiver(schema_needs_to_be_sync, sender=TenantMixin)
def created_user_client_in_background(sender, **kwargs):
    client = kwargs['tenant']
    print ("created_user_client_in_background %s" % client.schema_name)
    from clients.tasks import setup_tenant
    task = setup_tenant.delay(client)

@receiver(post_schema_sync, sender=TenantMixin)
def created_user_client(sender, **kwargs):

    client = kwargs['tenant']

    # send email to client to as tenant is ready to use

@receiver(schema_migrated, sender=run_migrations)
def handle_schema_migrated(sender, **kwargs):
    schema_name = kwargs['schema_name']

    # recreate materialized views in the schema

@receiver(schema_migrate_message, sender=run_migrations)
def handle_schema_migrate_message(**kwargs):
    message = kwargs['message']
    # recreate materialized views in the schema

Multi-types tenants

It is also possible to have different types of tenants. This is useful if you have two different types of users for instance you might want customers to use one style of tenant and suppliers to use another style. There is no limit to the amount of types however once the tenant has been set to a type it can’t easily be convert to another type. To enable multi types you need to change the setting file and add an extra field onto the tenant table.

In the setting file `SHARED_APPS`, `TENANT_APPS` and `PUBLIC_SCHEMA_URLCONF` needs to be removed.

The following needs to be added to the setting file

HAS_MULTI_TYPE_TENANTS = True
MULTI_TYPE_DATABASE_FIELD = 'type'  # or whatever the name you call the database field

TENANT_TYPES = {
    "public": {  # this is the name of the public schema from get_public_schema_name
        "APPS": ['django_tenants',
                 'django.contrib.admin',
                 'django.contrib.auth',
                 'django.contrib.contenttypes',
                 'django.contrib.sessions',
                 'django.contrib.messages',
                 'django.contrib.staticfiles',
                  # shared apps here
                  ],
        "URLCONF": "tenant_multi_types_tutorial.urls_public", # url for the public type here
    },
    "type1": {
        "APPS": ['django.contrib.contenttypes',
                 'django.contrib.auth',
                 'django.contrib.admin',
                 'django.contrib.sessions',
                 'django.contrib.messages',
                 # type1 apps here
                 ],
        "URLCONF": "tenant_multi_types_tutorial.urls_type1",
    },
    "type2": {
        "APPS": ['django.contrib.contenttypes',
                 'django.contrib.auth',
                 'django.contrib.admin',
                 'django.contrib.sessions',
                 'django.contrib.messages',
                 # type1 apps here
                 ],
        "URLCONF": "tenant_multi_types_tutorial.urls_type2",
    }
}

Now you need to change the install app line in the settings file

INSTALLED_APPS = []
for schema in TENANT_TYPES:
    INSTALLED_APPS += [app for app in TENANT_TYPES[schema]["APPS"] if app not in INSTALLED_APPS]

You also need to make sure that `ROOT_URLCONF` is blank

The tenant tables needs to have the following field added to the model

from django_tenants.utils import get_tenant_type_choices

class Client(TenantMixin):
    type = models.CharField(max_length=100, choices=get_tenant_type_choices())

That’s all you need to add the multiple types.

There is an example project called `tenant_multi_types`

Other settings

By default if no tenant is found it will raise an error Http404 however you add `SHOW_PUBLIC_IF_NO_TENANT_FOUND` to the setting it will display the the public tenant. This will not work for subfolders.

Admin

By default if you look at the admin all the tenant apps will be colored dark green you can disable this by doing.

TENANT_COLOR_ADMIN_APPS = False
Reverse

You can get the tenant domain name by calling a method on the tenant model called reverse.

Management commands

Every command except tenant_command runs by default on all tenants. You can also create your own commands that run on every tenant by inheriting BaseTenantCommand. To run only a particular schema, there is an optional argument called --schema.

Custom command example:

from django_tenants.management.commands import BaseTenantCommand
# rest of your imports

class Command(BaseTenantCommand):
    COMMAND_NAME = 'awesome command'
    # rest of your command
./manage.py migrate_schemas --schema=customer1
migrate_schemas

We’ve also packed the django migrate command in a compatible way with this app. It will also respect the SHARED_APPS and TENANT_APPS settings, so if you’re migrating the public schema it will only migrate SHARED_APPS. If you’re migrating tenants, it will only migrate TENANT_APPS.

./manage.py migrate_schemas

The options given to migrate_schemas are also passed to every migrate. Hence you may find handy

./manage.py migrate_schemas --list

Or

./manage.py migrate_schemas myapp 0001_initial --fake

in case you’re just switching your myapp application to use South migrations.

migrate_schemas in Parallel

You can run tenant migrations in parallel like this:

python manage.py migrate_schemas --executor=multiprocessing

In fact, you can write your own executor which will run tenant migrations in any way you want, just take a look at django_tenants/migration_executors.

The multiprocessing executor accepts the following settings:

  • TENANT_MULTIPROCESSING_MAX_PROCESSES (default: 2) - maximum number of processes for migration pool (this is to avoid exhausting the database connection pool)

  • TENANT_MULTIPROCESSING_CHUNKS (default: 2) - number of migrations to be sent at once to every worker

tenant_command

To run any command on an individual schema, you can use the special tenant_command, which creates a wrapper around your command so that it only runs on the schema you specify. For example

./manage.py tenant_command loaddata

If you don’t specify a schema, you will be prompted to enter one. Otherwise, you may specify a schema preemptively

./manage.py tenant_command loaddata --schema=customer1
all_tenants_command

To run any command on an every schema, you can use the special all_tenants_command, which creates a wrapper around your command so that it run on every schema. For example

./manage.py all_tenants_command loaddata
create_tenant_superuser

The command create_tenant_superuser is already automatically wrapped to have a schema flag. Create a new super user with

./manage.py create_tenant_superuser --username=admin --schema=customer1
create_tenant

The command create_tenant creates a new schema

./manage.py create_tenant --domain-domain=newtenant.net --schema_name=new_tenant --name=new_tenant --description="New tenant"

The argument are dynamic depending on the fields that are in the TenantMixin model. For example if you have a field in the TenantMixin model called company you will be able to set this using –company=MyCompany. If no argument are specified for a field then you be promted for the values. There is an additional argument of -s which sets up a superuser for that tenant.

delete_tenant

The command delete_tenant deletes a schema

./manage.py delete_tenant

Warning this command will delete a tenant and PostgreSQL schema regardless if auto_drop_schema is set to False.

clone_tenant

The command clone_tenant clones a schema.

./manage.py clone_tenant

There are some options to that can be set. You can view all the options by running

./manage.py clone_tenant -h

Credits to pg-clone-schema.

rename_schema

The command rename_schema renames a schema in the db and updates the Client associated with it.

./manage.py rename_schema

It will prompt you for the current name of the schema, and what it should be renamed to.

You can provide them with these arguments:

./manage.py rename_schema --rename_from old_name --rename_to new_name
create_missing_schemas

The command create_missing_schemas checks the tenant table against the list of schemas. If it find a schema that doesn’t exist it will create it.

./manage.py create_missing_schemas

PostGIS

If you want to run PostGIS add the following to your Django settings file

ORIGINAL_BACKEND = "django.contrib.gis.db.backends.postgis"

Performance Considerations

The hook for ensuring the search_path is set properly happens inside the DatabaseWrapper method _cursor(), which sets the path on every database operation. However, in a high volume environment, this can take considerable time. A flag, TENANT_LIMIT_SET_CALLS, is available to keep the number of calls to a minimum. The flag may be set in settings.py as follows:

#in settings.py:
TENANT_LIMIT_SET_CALLS = True

When set, django-tenants will set the search path only once per request. The default is False.

Extra Set Tenant Method

Sometime you might want to do something special when you switch to another schema / tenant such as read replica. Add EXTRA_SET_TENANT_METHOD_PATH to the settings file and point a method.

EXTRA_SET_TENANT_METHOD_PATH = 'tenant_multi_types_tutorial.set_tenant_utils.extra_set_tenant_stuff'
The method

The method takes 2 arguments the first is the database wrapper class and the second is the tenant. example

def extra_set_tenant_stuff(wrapper_class, tenant):
    pass

Logging

The optional TenantContextFilter can be included in settings.LOGGING to add the current schema_name and domain_url to the logging context.

# settings.py
LOGGING = {
    'filters': {
        'tenant_context': {
            '()': 'django_tenants.log.TenantContextFilter'
        },
    },
    'formatters': {
        'tenant_context': {
            'format': '[%(schema_name)s:%(domain_url)s] '
            '%(levelname)-7s %(asctime)s %(message)s',
        },
    },
    'handlers': {
        'console': {
            'filters': ['tenant_context'],
        },
    },
}

This will result in logging output that looks similar to:

[example:example.com] DEBUG 13:29 django.db.backends: (0.001) SELECT ...

Running in Development

If you want to use django-tenant in development you need to use a fake a domain name. All domains under the TLD .localhost will be routed to your local machine, so you can use things like tenant1.localhost and tenant2.localhost.

Migrating Single-Tenant to Multi-Tenant

Warning

The following instructions may or may not work for you. Use at your own risk!

  • Create a backup of your existing single-tenant database, presumably non PostgreSQL:

./manage.py dumpdata --all --indent 2 > database.json
  • Edit settings.py to connect to your new PostrgeSQL database

  • Execute manage.py migrate to create all tables in the PostgreSQL database

  • Ensure newly created tables are empty:

./manage.py sqlflush | ./manage.py dbshell
  • Load previously exported data into the database:

./manage.py loaddata --format json database.json
  • Create the public tenant:

./manage.py create_tenant

At this point your application should be multi-tenant aware and you may proceed creating more tenants.

Third Party Apps

Celery

Support for Celery is available at tenant-schemas-celery.

django-debug-toolbar

django-debug-toolbar routes need to be added to urls.py (both public and tenant) manually.

from django.conf import settings
from django.conf.urls import include

if settings.DEBUG:
    import debug_toolbar

    urlpatterns += patterns(
        '',
        url(r'^__debug__/', include(debug_toolbar.urls)),
    )

Useful information

Running code across every tenant

If you want to run some code on every tenant you can do the following

from django_tenants.utils import tenant_context, get_tenant_model

for tenant in get_tenant_model().objects.all():
    with tenant_context(tenant):
        pass
        # do whatever you want in that tenant

Examples

Tenant Tutorial

This app comes with an interactive tutorial to teach you how to use django-tenants and to demonstrate its capabilities. This example project is available under examples/tenant_tutorial. You will only need to edit the settings.py file to configure the DATABASES variable and then you’re ready to run

./manage.py runserver

All other steps will be explained by following the tutorial, just open http://127.0.0.1:8000 on your browser.

Running the example projects with Docker Compose

To run the example projects with docker-compose. You will need.

  1. Docker

  2. Docker Compose

Then you can run

docker-compose run web bash

cd examples/tenant_tutorial

python manage.py migrate

python manage.py create_tenant

python manage.py runserver 0.0.0.0:8088

All other steps will be explained by following the tutorial, just open http://127.0.0.1:8088 on your browser.

Tenant-aware file handling

The default Django behaviour is for all tenants to share one set of templates and static files between them. This can be changed so that each tenant will have its own:

  • Static files (like cascading stylesheets and JavaScript)

  • Location for files uploaded by users (usually stored in a /media directory)

  • Django templates

The process for making Django’s file handling tenant-aware generally consists of the following steps:

  1. Using a custom tenant-aware finder for locating files

  2. Specifying where finders should look for files

  3. Using a custom tenant-aware file storage handler for collecting and managing those files

  4. Using a custom tenant-aware loader for finding and loading Django templates

We’ll cover the configuration steps for each in turn.

Project layout

This configuration guide assumes the following Django project layout (loosely based on django-cookiecutter):

absolute/path/to/your_project_dir
    ...
    static              # System-wide static files
    templates           # System-wide templates
    # Tenant-specific files below will override pre-existing system-wide files with same name.
    tenants
        tenant_1        # Static files / templates for tenant_1
            templates
            static
        tenant_2        # Static files / templates for tenant_2
            templates
            static
    media               # Created automatically when users upload files
        tenant_1
        tenant_2
staticfiles             # Created automatically when collectstatic_schemas is run
    tenant_1
    tenant_2
    ...

The configuration details may differ depending on your specific requirements for your chosen layout. Fortunately, django-tenants makes it easy to cater for a wide range of project layouts as illustrated below.

Configuring the static file finders

Start by inserting django-tenants’ django_tenants.staticfiles.finders.TenantFileSystemFinder at the top of the list of available STATICFILES_FINDERS in your Django configuration file:

# in settings.py

STATICFILES_FINDERS = [
    "django_tenants.staticfiles.finders.TenantFileSystemFinder",  # Must be first
    "django.contrib.staticfiles.finders.FileSystemFinder",
    "django.contrib.staticfiles.finders.AppDirectoriesFinder",
    "compressor.finders.CompressorFinder",
]

# or this way

STATICFILES_FINDERS.insert(0, "django_tenants.staticfiles.finders.TenantFileSystemFinder")

By adding TenantFileSystemFinder at the top, we ensure that Django will look for the tenant-specific files first, before reverting to the standard search path. This makes it possible for tenants to override any static files (e.g. stylesheets or javascript files) that are specific to that tenant, and use the standard static files for the rest.

Next, add MULTITENANT_STATICFILES_DIRS to the configuration file in order to let TenantFileSystemFinder know where to look for tenant-specific static files:

# in settings.py

MULTITENANT_STATICFILES_DIRS = [
    os.path.join( "absolute/path/to/your_project_dir", "tenants/%s/static" ),
]

For the path provided above, %s will be replaced with the current tenant’s schema_name during runtime (see Specifying a different target directory for details).

Configuring the static files storage

By default, Django uses StaticFilesStorage for collecting static files into a dedicated folder on the server when the collectstatic management command is run. The location that the files are written to is specified in the STATIC_ROOT setting (usually configured to point to ‘staticfiles’).

django-tenants provides a replacement tenant-aware TenantStaticFilesStorage than can be configured by setting:

# in settings.py

STATICFILES_STORAGE = "django_tenants.staticfiles.storage.TenantStaticFilesStorage"

MULTITENANT_RELATIVE_STATIC_ROOT = ""  # (default: create sub-directory for each tenant)

The path specified in MULTITENANT_RELATIVE_STATIC_ROOT tells TenantStaticFilesStorage where in STATIC_ROOT the tenant’s files should be saved. The default behaviour is to just create a sub-directory for each tenant in STATIC_ROOT.

The command to collect the static files for all tenants is collectstatic_schemas. The optional --schema argument can be used to only collect files for a single tenant.

./manage.py collectstatic_schemas --schema=your_tenant_schema_name

Note

If you have configured an HTTP server, like nginx, to serve static files instead of the Django built-in server, then you also need to set REWRITE_STATIC_URLS = True. This tells django-tenants to rewrite STATIC_URL to include MULTITENANT_RELATIVE_STATIC_ROOT when static files are requested so that these files can be found and served directly by the external HTTP server.

Configuring media file storage

The default Django behavior is to store all files that are uploaded by users in one folder. The path for this upload folder can be configured via the standard MEDIA_ROOT setting.

The above behaviour can be changed for multi-tenant setups so that each tenant will have a dedicated sub-directory for storing user-uploaded files. To do this simply change DEFAULT_FILE_STORAGE so that TenantFileSystemStorage replaces the standard FileSystemStorage handler:

# in settings.py

DEFAULT_FILE_STORAGE = "django_tenants.files.storage.TenantFileSystemStorage"

MULTITENANT_RELATIVE_MEDIA_ROOT = ""  # (default: create sub-directory for each tenant)

The path specified in MULTITENANT_RELATIVE_MEDIA_ROOT tells TenantFileSystemStorage where in MEDIA_ROOT the tenant’s files should be saved. The default behaviour is to just create a sub-directory for each tenant in MEDIA_ROOT.

Configuring the template loaders

django-tenants provides a tenant-aware template loader that uses the current tenant’s schema_name when looking for templates.

It can be configured by inserting the custom Loader at the top of the list in the TEMPLATES setting, and specifying the template search path to be used in the MULTITENANT_TEMPLATE_DIRS setting, as illustrated below:

TEMPLATES = [
    {
        ...
        "DIRS": ["absolute/path/to/your_project_dir/templates"],  # -> Dirs used by the standard template loader
        "OPTIONS": {
            ...
            "loaders": [
                "django_tenants.template.loaders.filesystem.Loader",  # Must be first
                "django.template.loaders.filesystem.Loader",
                "django.template.loaders.app_directories.Loader",
            ],
            ...
        ...
    }
]

MULTITENANT_TEMPLATE_DIRS = [
    "absolute/path/to/your_project_dir/tenants/%s/templates"
]

Notice that django_tenants.template.loaders.filesystem.Loader is added at the top of the list. This will cause Django to look for the tenant-specific templates first, before reverting to the standard search path. This makes it possible for tenants to override individual templates as required.

Just like with standard Django, the first template found will be returned.

Attention

If the template contains any include tags, then all of the included templates need to be located in the tenant’s template folder as well. It is not currently possible to include templates from sources outside of the tenant’s template folder.

Specifying a different target directory

django-tenants supports simple Python string formatting for configuring the various path strings used throughout the configuration steps. any occurances of %s in the path string will be replaced with the current tenant’s schema_name during runtime.

This makes it possible to cater for more elaborate folder structures. Some examples are provided below:

# in settings.py

STATIC_ROOT = "absolute/path/to/your_project_dir/staticfiles"

MULTITENANT_RELATIVE_STATIC_ROOT = "tenants/%s

Static files will be collected into -> absolute/path/to/your_project_dir/staticfiles/tenants/schema_name.

…and for media files:

# in settings.py

MEDIA_ROOT = "absolute/path/to/your_project_dir/apps_dir/media/"

MULTITENANT_RELATIVE_MEDIA_ROOT = "%s/other_dir"

Media files will be uploaded at -> absolute/path/to/your_project_dir/apps_dir/media/schema_name/other_dir

Tests

Running the tests

Run these tests from the project dts_test_project, it comes prepacked with the correct settings file and extra apps to enable tests to ensure different apps can exist in SHARED_APPS and TENANT_APPS.

./manage.py test django_tenants.tests

If you want to run with custom migration executor then do

EXECUTOR=multiprocessing ./manage.py test django_tenants.tests

You can also run the tests with docker-compose

docker-compose run django-tenants-test

Updating your app’s tests to work with django_tenants

Because django will not create tenants for you during your tests, we have packed some custom test cases and other utilities. If you want a test to happen at any of the tenant’s domain, you can use the test case TenantTestCase. It will automatically create a tenant for you, set the connection’s schema to tenant’s schema and make it available at self.tenant. We have also included a TenantRequestFactory and a TenantClient so that your requests will all take place at the tenant’s domain automatically. Here’s an example

from django_tenants.test.cases import TenantTestCase
from django_tenants.test.client import TenantClient

class BaseSetup(TenantTestCase):

    def setUp(self):
        super().setUp()
        self.c = TenantClient(self.tenant)

    def test_user_profile_view(self):
        response = self.c.get(reverse('user_profile'))
        self.assertEqual(response.status_code, 200)

Additional information

You may have other fields on your tenant or domain model which are required fields. If you have there are two routines to look at setup_tenant and setup_domain

from django_tenants.test.cases import TenantTestCase
from django_tenants.test.client import TenantClient

class BaseSetup(TenantTestCase):

    @classmethod
    def setup_tenant(cls, tenant):
        """
        Add any additional setting to the tenant before it get saved. This is required if you have
        required fields.
        """
        tenant.required_value = "Value"
        return tenant

    def setup_domain(self, domain):
        """
        Add any additional setting to the domain before it get saved. This is required if you have
        required fields.
        """
        domain.ssl = True
        return domain

    def setUp(self):
        super().setUp()
        self.c = TenantClient(self.tenant)

    def test_user_profile_view(self):
        response = self.c.get(reverse('user_profile'))
        self.assertEqual(response.status_code, 200)

You can also change the test domain name and the test schema name by using get_test_schema_name and get_test_tenant_domain. by default the domain name is tenant.test.com and the schema name is test.

from django_tenants.test.cases import TenantTestCase
from django_tenants.test.client import TenantClient

class BaseSetup(TenantTestCase):
    @staticmethod
    def get_test_tenant_domain():
        return 'tenant.my_domain.com'


    @staticmethod
    def get_test_schema_name():
        return 'tester'

You can set the the verbosity by overriding the get_verbosity method.

Running tests faster

Using the TenantTestCase can make running your tests really slow quite early in your project. This is due to the fact that it drops, recreates the test schema and runs migrations for every TenantTestCase you have. If you want to gain speed, there’s a FastTenantTestCase where the test schema will be created and migrations ran only one time. The gain in speed is noticiable but be aware that by using this you will be perpertraiting state between your test cases, please make sure your they wont be affected by this.

Running tests using TenantTestCase can start being a bottleneck once the number of tests grow. If you do not care that the state between tests is kept, an alternative is to use the class FastTenantTestCase. Unlike TenantTestCase, the test schema and its migrations will only be created and ran once. This is a significant improvement in speed coming at the cost of shared state.

from django_tenants.test.cases import FastTenantTestCase

There are some extra method that you can use for FastTenantTestCase. They are.

flush_data default is True which means is will empty the table after each run. False will keep the data

class FastTenantTestCase(TenantTestCase):

    @classmethod
    def flush_data(cls):
        return True

use_existing_tenant Gets run if the setup doesn’t need to create a new database use_new_tenant Get run is an new database is created

class FastTenantTestCase(TenantTestCase):
    @classmethod
    def use_existing_tenant(cls):
        pass

    @classmethod
    def use_new_tenant(cls):
        pass

Get Involved!

Suggestions, bugs, ideas, patches, questions

Are highly welcome! Feel free to write an issue for any feedback you have or send a pull request on GitHub <https://github.com/django-tenants/django-tenants>. :)

Credits

django-tenant-schemas

I would like to thank the original author of this project Bernardo Pires Carneiro under the name django-tenant-schemas. I forked this project as I wanted to add enhancements to make it work better with Django 1.8

If you are using Django 1.7 or below please use his project.