Made a bunch of edits to docs/topics/cache.txt, mostly based on stuff from the Django Book

performance; yes, some cache types are faster than others.

Your cache preference goes in the ``CACHE_BACKEND`` setting in your settings

file. Here's an explanation of all available values for CACHE_BACKEND.

file. Here's an explanation of all available values for ``CACHE_BACKEND``.

Memcached

---------

By far the fastest, most efficient type of cache available to Django, Memcached

is an entirely memory-based cache framework originally developed to handle high

loads at LiveJournal.com and subsequently open-sourced by Danga Interactive.

It's used by sites such as Slashdot and Wikipedia to reduce database access and

It's used by sites such as Facebook and Wikipedia to reduce database access and

dramatically increase site performance.

Memcached is available for free at http://danga.com/memcached/ . It runs as a

daemon and is allotted a specified amount of RAM. All it does is provide an

interface -- a *lightning-fast* interface -- for adding, retrieving and

deleting arbitrary data in the cache. All data is stored directly in memory,

so there's no overhead of database or filesystem usage.

fast interface for adding, retrieving and deleting arbitrary data in the cache.

All data is stored directly in memory, so there's no overhead of database or

filesystem usage.

After installing Memcached itself, you'll need to install the Memcached Python

bindings. Two versions of this are available. Choose and install *one* of the

following modules:

bindings, which are not bundled with Django directly. Two versions of this are

available. Choose and install *one* of the following modules:

    * The fastest available option is a module called ``cmemcache``, available

      at http://gijsbert.org/cmemcache/ .

    CACHE_BACKEND = 'memcached://127.0.0.1:11211/'

One excellent feature of Memcached is its ability to share cache over multiple

servers. To take advantage of this feature, include all server addresses in

``CACHE_BACKEND``, separated by semicolons. In this example, the cache is

shared over Memcached instances running on IP address 172.19.26.240 and

172.19.26.242, both on port 11211::

servers. This means you can run Memcached daemons on multiple machines, and the

program will treat the group of machines as a *single* cache, without the need

to duplicate cache values on each machine. To take advantage of this feature,

include all server addresses in ``CACHE_BACKEND``, separated by semicolons.

In this example, the cache is shared over Memcached instances running on IP

address 172.19.26.240 and 172.19.26.242, both on port 11211::

    CACHE_BACKEND = 'memcached://172.19.26.240:11211;172.19.26.242:11211/'

Memory-based caching has one disadvantage: Because the cached data is stored in

memory, the data will be lost if your server crashes. Clearly, memory isn't

intended for permanent data storage, so don't rely on memory-based caching as

your only data storage. Actually, none of the Django caching backends should be

used for permanent storage -- they're all intended to be solutions for caching,

not storage -- but we point this out here because memory-based caching is

In the following example, the cache is shared over Memcached instances running

on the IP addresses 172.19.26.240 (port 11211), 172.19.26.242 (port 11212), and

172.19.26.244 (port 11213)::

    CACHE_BACKEND = 'memcached://172.19.26.240:11211;172.19.26.242:11212;172.19.26.244:11213/'

A final point about Memcached is that memory-based caching has one

disadvantage: Because the cached data is stored in memory, the data will be

lost if your server crashes. Clearly, memory isn't intended for permanent data

storage, so don't rely on memory-based caching as your only data storage.

Without a doubt, *none* of the Django caching backends should be used for

permanent storage -- they're all intended to be solutions for caching, not

storage -- but we point this out here because memory-based caching is

particularly temporary.

Database caching

    CACHE_BACKEND = 'db://my_cache_table'

The database caching backend uses the same database as specified in your

settings file. You can't use a different database backend for your cache table.

Database caching works best if you've got a fast, well-indexed database server.

Filesystem caching

Note that there are three forward slashes toward the beginning of that example.

The first two are for ``file://``, and the third is the first character of the

directory path, ``/var/tmp/django_cache``.

directory path, ``/var/tmp/django_cache``. If you're on Windows, put the

drive letter after the ``file://``, like this::

    file://c:/foo/bar

The directory path should be absolute -- that is, it should start at the root

of your filesystem. It doesn't matter whether you put a slash at the end of the

directory ``/var/tmp/django_cache`` exists and is readable and writable by the

user ``apache``.

Each cache value will be stored as a separate file whose contents are the

cache data saved in a serialized ("pickled") format, using Python's ``pickle``

module. Each file's name is the cache key, escaped for safe filesystem use.

Local-memory caching

--------------------

Note that each process will have its own private cache instance, which means no

cross-process caching is possible. This obviously also means the local memory

cache isn't particularly memory-efficient, so it's probably not a good choice

for production environments.

for production environments. It's nice for development.

Dummy caching (for development)

-------------------------------

just implements the cache interface without doing anything.

This is useful if you have a production site that uses heavy-duty caching in

various places but a development/test environment on which you don't want to

cache. As a result, your development environment won't use caching and your

production environment still will. To activate dummy caching, set

``CACHE_BACKEND`` like so::

various places but a development/test environment where you don't want to cache

and don't want to have to change your code to special-case the latter. To

activate dummy caching, set ``CACHE_BACKEND`` like so::

    CACHE_BACKEND = 'dummy:///'

CACHE_BACKEND arguments

-----------------------

All caches may take arguments. They're given in query-string style on the

``CACHE_BACKEND`` setting. Valid arguments are:

Each cache backend may take arguments. They're given in query-string style on

the ``CACHE_BACKEND`` setting. Valid arguments are as follows:

    timeout

        Default timeout, in seconds, to use for the cache. Defaults to 5

        minutes (300 seconds).

    * ``timeout``: The default timeout, in seconds, to use for the cache.

      This argument defaults to 300 seconds (5 minutes).

    max_entries

        For the ``locmem``, ``filesystem`` and ``database`` backends, the

        maximum number of entries allowed in the cache before it is cleaned.

        Defaults to 300.

    * ``max_entries``: For the ``locmem``, ``filesystem`` and ``database``

      backends, the maximum number of entries allowed in the cache before old

      values are deleted. This argument defaults to 300.

    cull_percentage

        The percentage of entries that are culled when max_entries is reached.

        The actual percentage is 1/cull_percentage, so set cull_percentage=3 to

        cull 1/3 of the entries when max_entries is reached.

    * ``cull_percentage``: The percentage of entries that are culled when

      ``max_entries`` is reached. The actual ratio is ``1/cull_percentage``, so

      set ``cull_percentage=2`` to cull half of the entries when ``max_entries``

      is reached.

        A value of 0 for cull_percentage means that the entire cache will be

        dumped when max_entries is reached. This makes culling *much* faster

        at the expense of more cache misses.

      A value of ``0`` for ``cull_percentage`` means that the entire cache will

      be dumped when ``max_entries`` is reached. This makes culling *much*

      faster at the expense of more cache misses.

In this example, ``timeout`` is set to ``60``::

    * Sets the ``Last-Modified`` header to the current date/time when a fresh

      (uncached) version of the page is requested.

    * Sets the ``Expires`` header to the current date/time plus the defined

      ``CACHE_MIDDLEWARE_SECONDS``.

* Sets the ``Cache-Control`` header to give a max age for the page -- again,

  from the ``CACHE_MIDDLEWARE_SECONDS`` setting.

    * Sets the ``Cache-Control`` header to give a max age for the page --

      again, from the ``CACHE_MIDDLEWARE_SECONDS`` setting.

See :ref:`topics-http-middleware` for more on middleware.

    from django.views.decorators.cache import cache_page

    def slashdot_this(request):

    def my_view(request):

        ...

    slashdot_this = cache_page(slashdot_this, 60 * 15)

    my_view = cache_page(my_view, 60 * 15)

Or, using Python 2.4's decorator syntax::

    @cache_page(60 * 15)

    def slashdot_this(request):

    def my_view(request):

        ...

``cache_page`` takes a single argument: the cache timeout, in seconds. In the

above example, the result of the ``slashdot_this()`` view will be cached for 15

minutes.

above example, the result of the ``my_view()`` view will be cached for 15

minutes. (Note that we've written it as ``60 * 15`` for the purpose of

readability. ``60 * 15`` will be evaluated to ``900`` -- that is, 15 minutes

multiplied by 60 seconds per minute.)

The per-view cache, like the per-site cache, is keyed off of the URL. If

multiple URLs point at the same view, each URL will be cached separately.

Continuing the ``my_view`` example, if your URLconf looks like this::

    urlpatterns = ('',

        (r'^foo/(\d{1,2})/$', my_view),

    )

then requests to ``/foo/1/`` and ``/foo/23/`` will be cached separately, as

you may expect. But once a particular URL (e.g., ``/foo/23/``) has been

requested, subsequent requests to that URL will use the cache.

Specifying per-view cache in the URLconf

----------------------------------------

The examples in the previous section have hard-coded the fact that the view is

cached, because ``cache_page`` alters the ``my_view`` function in place. This

approach couples your view to the cache system, which is not ideal for several

reasons. For instance, you might want to reuse the view functions on another,

cache-less site, or you might want to distribute the views to people who might

want to use them without being cached. The solution to these problems is to

specify the per-view cache in the URLconf rather than next to the view functions

themselves.

Doing so is easy: simply wrap the view function with ``cache_page`` when you

refer to it in the URLconf. Here's the old URLconf from earlier::

    urlpatterns = ('',

        (r'^foo/(\d{1,2})/$', my_view),

    )

Here's the same thing, with ``my_view`` wrapped in ``cache_page``::

    from django.views.decorators.cache import cache_page

    urlpatterns = ('',

        (r'^foo/(\d{1,2})/$', cache_page(my_view, 60 * 15)),

    )

If you take this approach, don't forget to import ``cache_page`` within your

URLconf.

Template fragment caching

=========================

The low-level cache API

=======================

Sometimes, however, caching an entire rendered page doesn't gain you very much.

For example, you may find it's only necessary to cache the result of an

intensive database query. In cases like this, you can use the low-level cache

API to store objects in the cache with any level of granularity you like.

Sometimes, caching an entire rendered page doesn't gain you very much and is,

in fact, inconvenient overkill.

Perhaps, for instance, your site includes a view whose results depend on 

several expensive queries, the results of which change at different intervals.

In this case, it would not be ideal to use the full-page caching that the 

per-site or per-view cache strategies offer, because you wouldn't want to 

cache the entire result (since some of the data changes often), but you'd still 

want to cache the results that rarely change.

For cases like this, Django exposes a simple, low-level cache API. You can use

this API to store objects in the cache with any level of granularity you like.

You can cache any Python object that can be pickled safely: strings,

dictionaries, lists of model objects, and so forth. (Most common Python objects

can be pickled; refer to the Python documentation for more information about

pickling.)

The cache API is simple. The cache module, ``django.core.cache``, exports a

``cache`` object that's automatically created from the ``CACHE_BACKEND``

setting::

The cache module, ``django.core.cache``, has a ``cache`` object that's

automatically created from the ``CACHE_BACKEND`` setting::

    >>> from django.core.cache import cache

If the object doesn't exist in the cache, ``cache.get()`` returns ``None``::

    >>> cache.get('some_other_key')

    None

    # Wait 30 seconds for 'my_key' to expire...

    >>> cache.get('my_key')

    None

get() can take a ``default`` argument::

We advise against storing the literal value ``None`` in the cache, because you

won't be able to distinguish between your stored ``None`` value and a cache

miss signified by a return value of ``None``.

``cache.get()`` can take a ``default`` argument. This specifies which value to

return if the object doesn't exist in the cache::

    >>> cache.get('my_key', 'has expired')

    'has expired'

    backends that support atomic increment/decrement (most notably, the

    memcached backend), increment and decrement operations will be atomic.

    However, if the backend doesn't natively provide an increment/decrement

    operation, it will be implemented using a 2 step retrieve/update.

That's it. The cache has very few restrictions: You can cache any object that

can be pickled safely, although keys must be strings.

    operation, it will be implemented using a two-step retrieve/update.

Upstream caches

===============

Here are a few examples of upstream caches:

    * Your ISP may cache certain pages, so if you requested a page from

      somedomain.com, your ISP would send you the page without having to access

      somedomain.com directly.

      http://example.com/, your ISP would send you the page without having to

      access example.com directly. The maintainers of example.com have no

      knowledge of this caching; the ISP sits between example.com and your Web

      browser, handling all of the caching transparently.

    * Your Django Web site may sit behind a Squid Web proxy

      (http://www.squid-cache.org/) that caches pages for performance. In this

      case, each request first would be handled by Squid, and it'd only be

      passed to your application if needed.

    * Your Django Web site may sit behind a *proxy cache*, such as Squid Web

      Proxy Cache (http://www.squid-cache.org/), that caches pages for

      performance. In this case, each request first would be handled by the

      proxy, and it would be passed to your application only if needed.

    * Your Web browser caches pages, too. If a Web page sends out the right

      headers, your browser will use the local (cached) copy for subsequent

      requests to that page.

    * Your Web browser caches pages, too. If a Web page sends out the

      appropriate headers, your browser will use the local cached copy for

      subsequent requests to that page, without even contacting the Web page

      again to see whether it has changed.

Upstream caching is a nice efficiency boost, but there's a danger to it:

Many Web pages' contents differ based on authentication and a host of other

his user-specific inbox page cached for subsequent visitors to the site. That's

not cool.

Fortunately, HTTP provides a solution to this problem: A set of HTTP headers

exist to instruct caching mechanisms to differ their cache contents depending

on designated variables, and to tell caching mechanisms not to cache particular

pages.

Fortunately, HTTP provides a solution to this problem. A number of HTTP headers

exist to instruct upstream caches to differ their cache contents depending on

designated variables, and to tell caching mechanisms not to cache particular

pages. We'll look at some of these headers in the sections that follow.

Using Vary headers

==================

One of these headers is ``Vary``. It defines which request headers a cache

The ``Vary`` header defines which request headers a cache

mechanism should take into account when building its cache key. For example, if

the contents of a Web page depend on a user's language preference, the page is

said to "vary on language."

By default, Django's cache system creates its cache keys using the requested

path -- e.g., ``"/stories/2005/jun/23/bank_robbed/"``. This means every request

path (e.g., ``"/stories/2005/jun/23/bank_robbed/"``). This means every request

to that URL will use the same cached version, regardless of user-agent

differences such as cookies or language preferences.

That's where ``Vary`` comes in.

If your Django-powered page outputs different content based on some difference

in request headers -- such as a cookie, or language, or user-agent -- you'll

need to use the ``Vary`` header to tell caching mechanisms that the page output

depends on those things.

differences such as cookies or language preferences. However, if this page

produces different content based on some difference in request headers -- such

as a cookie, or a language, or a user-agent -- you'll need to use the ``Vary``

header to tell caching mechanisms that the page output depends on those things.

To do this in Django, use the convenient ``vary_on_headers`` view decorator,

like so::

    # Python 2.3 syntax.

    def my_view(request):

        ...

        # ...

    my_view = vary_on_headers(my_view, 'User-Agent')

    # Python 2.4 decorator syntax.

    # Python 2.4+ decorator syntax.

    @vary_on_headers('User-Agent')

    def my_view(request):

        ...

        # ...

In this case, a caching mechanism (such as Django's own cache middleware) will

cache a separate version of the page for each unique user-agent.

The advantage to using the ``vary_on_headers`` decorator rather than manually

setting the ``Vary`` header (using something like

``response['Vary'] = 'user-agent'``) is that the decorator adds to the ``Vary``

header (which may already exist) rather than setting it from scratch.

``response['Vary'] = 'user-agent'``) is that the decorator *adds* to the

``Vary`` header (which may already exist), rather than setting it from scratch

and potentially overriding anything that was already in there.

You can pass multiple headers to ``vary_on_headers()``::

    @vary_on_headers('User-Agent', 'Cookie')

    def my_view(request):

        ...

        # ...

Because varying on cookie is such a common case, there's a ``vary_on_cookie``

This tells upstream caches to vary on *both*, which means each combination of

user-agent and cookie will get its own cache value. For example, a request with

the user-agent ``Mozilla`` and the cookie value ``foo=bar`` will be considered

different from a request with the user-agent ``Mozilla`` and the cookie value

``foo=ham``.

Because varying on cookie is so common, there's a ``vary_on_cookie``

decorator. These two views are equivalent::

    @vary_on_cookie

    def my_view(request):

        ...

        # ...

    @vary_on_headers('Cookie')

    def my_view(request):

        ...

        # ...

Also note that the headers you pass to ``vary_on_headers`` are not case

sensitive. ``"User-Agent"`` is the same thing as ``"user-agent"``.

The headers you pass to ``vary_on_headers`` are not case sensitive;

``"User-Agent"`` is the same thing as ``"user-agent"``.

You can also use a helper function, ``django.utils.cache.patch_vary_headers``,

directly::

directly. This function sets, or adds to, the ``Vary header``. For example::

    from django.utils.cache import patch_vary_headers

    def my_view(request):

        ...

        # ...

        response = render_to_response('template_name', context)

        patch_vary_headers(response, ['Cookie'])

        return response

``patch_vary_headers`` takes an ``HttpResponse`` instance as its first argument

and a list/tuple of header names as its second argument.

and a list/tuple of case-insensitive header names as its second argument.

For more on Vary headers, see the `official Vary spec`_.

Controlling cache: Using other headers

======================================

Another problem with caching is the privacy of data and the question of where

Other problems with caching are the privacy of data and the question of where

data should be stored in a cascade of caches.

A user usually faces two kinds of caches: his own browser cache (a private

cache) and his provider's cache (a public cache). A public cache is used by

multiple users and controlled by someone else. This poses problems with

sensitive data: You don't want, say, your banking-account number stored in a

A user usually faces two kinds of caches: his or her own browser cache (a

private cache) and his or her provider's cache (a public cache). A public cache

is used by multiple users and controlled by someone else. This poses problems

with sensitive data--you don't want, say, your bank account number stored in a

public cache. So Web applications need a way to tell caches which data is

private and which is public.

Django, use the ``cache_control`` view decorator. Example::

    from django.views.decorators.cache import cache_control

    @cache_control(private=True)

    def my_view(request):

        ...

        # ...

This decorator takes care of sending out the appropriate HTTP header behind the

scenes.

allows applications to do the following:

    * Define the maximum time a page should be cached.

    * Specify whether a cache should always check for newer versions, only

      delivering the cached content when there are no changes. (Some caches

      might deliver cached content even if the server page changed -- simply

      might deliver cached content even if the server page changed, simply

      because the cache copy isn't yet expired.)

In Django, use the ``cache_control`` view decorator to specify these cache

parameters. In this example, ``cache_control`` tells caches to revalidate the

cache on every access and to store cached versions for, at most, 3600 seconds::

cache on every access and to store cached versions for, at most, 3,600 seconds::

    from django.views.decorators.cache import cache_control

    @cache_control(must_revalidate=True, max_age=3600)

    def my_view(request):

        ...

        # ...

Any valid ``Cache-Control`` HTTP directive is valid in ``cache_control()``.

Here's a full list:

If you want to use headers to disable caching altogether,

``django.views.decorators.cache.never_cache`` is a view decorator that adds

headers to ensure the response won't be cached by browsers or other caches. Example::

headers to ensure the response won't be cached by browsers or other caches.

Example::

    from django.views.decorators.cache import never_cache

    @never_cache

    def myview(request):

        ...

        # ...

.. _`Cache-Control spec`: http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.9

apps' performance:

    * ``django.middleware.http.ConditionalGetMiddleware`` adds support for

      conditional GET. This makes use of ``ETag`` and ``Last-Modified``

      headers.

      modern browsers to conditionally GET responses based on the ``ETag`` 

      and ``Last-Modified`` headers.

    * ``django.middleware.gzip.GZipMiddleware`` compresses content for browsers

      that understand gzip compression (all modern browsers).

    * ``django.middleware.gzip.GZipMiddleware`` compresses responses for all 

      moderns browsers, saving bandwidth and transfer time.

Order of MIDDLEWARE_CLASSES

===========================