Full featured redis cache backend for Django.
NOTE: The 3.0 version has many changes in the code, and may have regressiones. If you want a version more stable and proven, use the 2.x branch
Redis cache allows you to use either a TCP connection or Unix domain socket to connect to your redis server. Using a TCP connection is useful for when you have your redis server separate from your app server and/or within a distributed environment. Unix domain sockets are useful if you have your redis server and application running on the same machine and want the fastest possible connection.
You can specify (optionally) what parser class you want redis-py to use
when parsing messages from the redis server. redis-py will pick the best
parser for you implicitly, but using the PARSER_CLASS
setting gives you
control and the option to roll your own parser class if you are so bold.
This cache backend requires the redis-py python client library for communicating with the Redis server.
- In active development.
- Sharding supported in a single package.
- Complete battery of tests (accepting improvements).
- Used in production environments on several projects.
- Can set keys with infinite timeout:
cache.set('key', 'value', timeout=0)
- Pluggable clients.
- Python3 support with same codebase.
- Same behavior as memcached backend with connection exceptions.
- Supports Django: 1.3, 1.4 and 1.5
- You can take advantage of the connection pool to directly access to the connection object of redis.
- Auto failover
- Master-Slave pluggable client.
Currently, for django versions >1.3
, migration is very easy, since there is no significant difference in connection APIs.
The only point is the connection string that since version 3.0 has changed slightly. You can look at the differences in the later sections.
Run python setup.py install
to install,
or place redis_cache
on your Python path.
You can also install it with: pip install django-redis
To start using django-redis
, you must change your django cache settings. django-redis
implements the standard interface for django cache backends.
With django-redis==3.0
has introduced certain backwards incompatible changes, in part of redis connection settings (connection string).
The LOCATION
attr string must be have always three colons instead of two.
Old way (django-redis < 3.0):
CACHES = {
#...
"LOCATION": "ip:port",
"OPTIONS": {
"DB": 1
}
}
New way:
CACHES = {
#...
"LOCATION": "ip:port:db",
}
This is the complete example using a tcp connection:
CACHES = {
"BACKEND": "redis_cache.cache.RedisCache",
"LOCATION": "127.0.0.1:6379:1",
"OPTIONS": {
"CLIENT_CLASS": "redis_cache.client.DefaultClient",
}
}
And this is a complete example using unix sockets:
# When using unix domain sockets
# Note: ``LOCATION`` needs to be the same as the ``unixsocket`` setting
# in your redis.conf
CACHES = {
'default': {
'BACKEND': 'redis_cache.cache.RedisCache',
'LOCATION': 'unix:/path/to/socket/file.sock:1',
'OPTIONS': {
'PASSWORD': 'foopassword',
'PICKLE_VERSION': -1, # default
'PARSER_CLASS': 'redis.connection.HiredisParser'
},
},
}
Optionally, with PARSER_CLASS="redis.connection.HiredisParser"
you can set hiredis parser.
The configuration is same as a default with unique diference: the LOCATION
attr must
be a list of connection strings.
Some example:
CACHES = {
"LOCATION": [
"127.0.0.1:6379:1",
"127.0.0.1:6379:2",
],
"OPTIONS": {
"CLIENT_CLASS": "redis_cache.client.ShardClient",
}
}
django-redis
provides 2 additional methods to the standard django-cache api interface:
cache.keys(wildcard_pattern)
- Add abilite to retrieve a list of keys with wildcard pattern.cache.delete_pattern(wildcard_pattern)
- Same askeys
, but this delete all keys matching the wildcard pattern.
Example:
from django.core.cache import cache
# this returns all keys starts with ``session_``
result = cache.keys("session_*")
# delete all keys stats with ``session_``
cache.delete_pattern("session_*")
And sometimes, our application requires direct access to redis, besides the standard cache.
Instead of repeating the code 2 times and create multiple connection pool, django-redis
exposes a simple API to access
the redis client directly, bypassing the cache API. This allows an application that needs the cache API and direct access to redis,
have everything in one.
Example:
>>> from redis_cache import get_redis_connection
>>> con = get_redis_connection('default')
>>> con
<redis.client.Redis object at 0x2dc4510>
NOTE: not all pluggable clients supports this feature. The simple example is a ShardClient, this does not supports access to raw redis connection.