DRF Cache provides easy to use, powerful and flexible cache framework for django-rest-framwork apps.
Install using pip
...
pip install rest-framework-cache
Add 'rest_framework_cache'
to your INSTALLED_APPS
setting.
INSTALLED_APPS = (
...
'rest_framework_cache',
)
You must execute autodiscover to load your serializers. To do this change your urls.py
adding the following code (sunch as Django admin):
from rest_framework_cache.registry import cache_registry
cache_registry.autodiscover()
This lib does not install any dependency, but your project obviously have to be using Django Rest Framework.
To use the DRF cache you must register your serializer into cache registry ( such as Django admin ). You also must change your serializer to inherit the CachedSerializerMixin
:
from rest_framework import serializers
# You must import the CachedSerializerMixin and cache_registry
from rest_framework_cache.serializers import CachedSerializerMixin
from rest_framework_cache.registry import cache_registry
from .models import Comment
class CommentSerializer(serializers.ModelSerializer, CachedSerializerMixin):
class Meta:
model = Comment
cache_registry.register(CommentSerializer)
To the cache successfully work you must configure the Django CACHES setting. We recomend that you take a look on Django cache docs here https://docs.djangoproject.com/en/1.9/topics/cache/
If you need use a cache backend different of the default you can specify it on the RF_CACHE_BACKEND
.
To do this edit your settings.py
like this:
# ...
CACHES = {
'default': {
'BACKEND': 'django.core.cache.backends.memcached.MemcachedCache',
'LOCATION': '127.0.0.1:11211',
},
'rest_backend': {
'BACKEND': 'django.core.cache.backends.locmem.LocMemCache',
'LOCATION': 'unique-snowflake',
}
}
REST_FRAMEWORK_CACHE = {
'DEFAULT_CACHE_BACKEND': 'rest_backend',
}
# ...
You can set the cache timeout using DEFAULT_CACHE_TIMEOUT
.
REST_FRAMEWORK_CACHE = {
'DEFAULT_CACHE_TIMEOUT': 86400, # Default is 1 day
}
When the representation of CachedSerializerMixin
is called the fist thing that will be executed is a verification that checks if the request objects is already in cache, if yes the cached object will be returned without touch the database, otherwise the object will be requested to the database stored on cache and returned.
When your serializer is declareted using the CachedSerializerMixin
the DRF cache register a signal to the serializer model. When a instance of the model has changed or deleted the signal clear related object on the cache backend.