cookbook_django.txt

== link:index.html[Index] -> link:cookbook.html[Cookbook]
///////////////////////////////////////////////////////////////////
Last checked:
* Cherokee: 1.2.2
* Django 1.3
* 2011/03/23
///////////////////////////////////////////////////////////////////

Cookbook: Setting up Django
---------------------------

Django is a fantastic high-level Python Web framework that can run
nicely with Cherokee and either the
link:modules_handlers_scgi.html[SCGI] or
link:modules_handlers_fcgi.html[FastCGI] handlers.

To properly set up Cherokee to use Django you will need a working
Django framework, which is not difficult at all to get ready. The
details vary from system to system. You can refer to the official
project's documentation for more information on
link:http://www.djangoproject.com/documentation/install[how to install
Django].

There are two ways to deploy a Django project running on Cherokee. The
first one is the classical method and involves the usage of Flup, it
is the method documented in this recipe and has been the standard
method for years. Thousands of sites currently use this approach
successfully.

The other method involves using link:cookbook_uwsgi.html[uWSGI]. It is
a new, modern and feature-rich approach. There is a separate recipe
and wizard for that, so check it out. We can assure you it will be
worth your time. The instructions on how to set up the Framework are
the same, but the method in which Cherokee is configured varies.

There is a
link:http://www.cherokee-project.com/screencasts.html#django_flup[screencast]
available at the
link:http://www.cherokee-project.com/[Cherokee-Project website] to
demonstrate how easy it is to use the Django wizard.

.Django wizard
image:media/images/screencast.png["Django wizard",
link="http://www.cherokee-project.com/screencasts.html#django_flup"]


=== Preparing the framework

On Debian based systems this will be enough:
----
# apt-get install python-django python-flup
----

You will need Flup because it implements the standard interface
between Python Web applications and Web servers, so you will be using
it to run your web application either as FastCGI or SCGI.

Once you are done with that, you must deploy your Django project:

----
$ cd /var/www
$ django-admin startproject example
----

This will create the basic structure into a new directory called
`example`.

It is important to ensure that `manage.py` has execution permissions.
If it does not, you will only have to set them by running:

----
chmod a+rx manage.py
----

Now you are ready to configure Cherokee.

=== Preparing Cherokee

You can either do it by hand, or you can use the appropriate wizard
for a hassle-free configuration.

To do so, simply access the `Virtual Servers` section in
`Cherokee-Admin`, and click on the `Add` button at the top of the
panel to show the list of available wizards. `Django` is located under
the `Platforms` category. You can use this wizard both within your
`Virtual Servers` list (in which case a new virtual server is
configured), and from the `Rule Management` button on the `Behavior`
tab inside one of your virtual servers. In this last scenario, the
Django project will be accessible under the specified web directory of
the selected virtual server.

.Wizards
image::media/images/admin_vservers_wizard.png[Virtual Server Wizards]

Simply fill in the required fields. `Project Directory` applies to the
path where your Django application is located, while `Document Root`
applies to the directory that contains non-Django related files, such
as any static content.

That's it. You can stop reading now. Go enjoy your Django application
running on Cherokee.

=== Setting up Cherokee by hand

The following information details the steps needed to configure
Cherokee by hand, which is not the recommended option. The wizards are
constantly being updated as new releases of the software become
available, while the recipe might become somehow more outdated given
enough time.

It is not difficult, though. You only need to know how to spawn the
FastCGI or SCGI, which is done with a script provided by your project
called `manage.py`.

Django can be run on a TCP port or on a Unix socket. In our example we
will be launching it as threaded server on a TCP port with SCGI
protocol. This is accomplished with the following command, which is
what we will have to set up in `cherokee-admin`.

----
./manage.py runfcgi method=threaded host=127.0.0.1 port=3033 protocol=scgi
----

The process is fairly simple. Set up a new rule for this new path and
manage it with the SCGI handler. If you wanted to use the FastCGI
handler instead you would only have to omit the last parameter and
FastCGI would be used by default. The configuration of the handler is
exactly the same for SCGI and FastCGI.

Once you have created the new rule for your `/var/www/example`
directory, choose the desired handler and use the following
configuration.

.Common CGI options
Under `Common CGI options` make sure to check the `Error handler` box and
uncheck `Check file`. This is to prevent possible errors with the
`INFO_PATH` generation that can happen when an application, in this
case 'Django', manages the whole subtree. This is mentioned in the
link:modules_handlers_cgi.html[Common CGI] section of the
documentation. It is a good idea to enable the `Error handler`
checkbox since it will help you determine if an error is associated
with your Django application or with Cherokee. This, however, is not
required.

.SCGI handler
image::media/images/cookbook_django_common.png[Common CGI options]

.SCGI specific
Under `SCGI specific` make sure to add the hosts providing the
service. This is done by adding one or more information sources.

.Django information source
image::media/images/cookbook_django_infosources.png[Information Sources]

Note that you will have to manually launch the `spawner` if
you use a `Remote host` as `Information source` instead of a `Local
interpreter`.

You will simply have to add as many sources as needed, for instance
our example uses one nicknamed `django1`, created as *local
interpreter* with these parameters on port 3033.

[cols="10%,90%",options="header"]
|======================================================================
|Host           |Interpreter
|localhost:3033 |`/var/www/example/manage.py runfcgi method=threaded ` \
                 `host=127.0.0.1 port=3033 protocol=scgi`
|======================================================================

You can set up as many hosts as desired and Cherokee will balance the
load among them.

Once everything is done you can check if Django is really
working. Just navigate to the path configured by your rule,
http://localhost/example for instance, and you should see some notes
about your recently created project.

.It worked!
image::media/images/cookbook_django.png[Django example]

*********************************************************************
Should you wish to deploy your Django application directly instead of
through `manage.py` this could be easily achieved. Just remember to
add the appropriate command used to launch the interpreter in the
relevant `information source`, and it will be executed. As noted in
the warning on the documentation about
link:config_info_sources.html#daemonization[Information Sources], Cherokee
is smart enough to daemonize the backend during the spawning process,
so you must avoid any manual daemonization on your part to prevent the
processess from interfering.

This is done adding the appropriate parameter to the `runfastcgi`
method of `django.core.servers.fastcgi` in your script.

.Using a unix socket:
----
from django.core.servers.fastcgi import runfastcgi

runfastcgi(method="threaded", daemonize="false", maxrequests=5,
           protocol="scgi", socket="/tmp/cherokee-django.socket",
           pidfile="/tmp/cherokee-django.pid")
----

.Using a host:port configuration:
----
from django.core.servers.fastcgi import runfastcgi

runfastcgi(method="threaded", daemonize="false", maxrequests=5,
           protocol="scgi", host="127.0.0.1", port=3033)
----
*********************************************************************