Django Async Redis

Introduction

django-async-redis is a full featured Redis cache and session backend for Django.

• Free software: Apache Software License 2.0
• Documentation: https://django-async-redis.readthedocs.io.

Requirements

• Python 3.7+
• Django 3.2+
• redis-py 4.2+
• Redis server 2.8+

User guide

Installation

Install with pip:

$ python -m pip install django-async-redis

Configure as cache backend

To start using django-async-redis, you should change your Django cache settings to something like:

CACHES = {
    "default": {
        "BACKEND": "django_async_redis.cache.RedisCache",
        "LOCATION": "redis://127.0.0.1:6379/1",
        "OPTIONS": {
            "CLIENT_CLASS": "django_async_redis.client.DefaultClient",
        }
    }
}

django-async-redis uses the aioredis native URL notation for connection strings, it allows better interoperability and has a connection string in more "standard" way. Some examples:

• redis://[:password]@localhost:6379/0
• rediss://[:password]@localhost:6379/0
• unix://[:password]@/path/to/socket.sock?db=0

Three URL schemes are supported:

• redis://: creates a normal TCP socket connection
• rediss://: creates a SSL wrapped TCP socket connection
• unix:// creates a Unix Domain Socket connection

There are several ways to specify a database number:

• A db querystring option, e.g. redis://localhost?db=0
• If using the redis:// scheme, the path argument of the URL, e.g. redis://localhost/0

Advanced usage

Pickle version

For almost all values, django-async-redis uses pickle to serialize objects.

The latest available version of pickle is used by default. If you want set a concrete version, you can do it, using PICKLE_VERSION option:

CACHES = {
    "default": {
        # ...
        "OPTIONS": {
            "PICKLE_VERSION": -1  # Use the latest protocol version
        }
    }
}

Memcached exceptions behavior

In some situations, when Redis is only used for cache, you do not want exceptions when Redis is down. This is default behavior in the memcached backend and it can be emulated in django-async-redis.

For setup memcached like behaviour (ignore connection exceptions), you should set IGNORE_EXCEPTIONS settings on your cache configuration:

CACHES = {
    "default": {
        # ...
        "OPTIONS": {
            "IGNORE_EXCEPTIONS": True,
        }
    }
}

Also, you can apply the same settings to all configured caches, you can set the global flag in your settings:

DJANGO_ASYNC_REDIS_IGNORE_EXCEPTIONS = True

Log Ignored Exceptions

When ignoring exceptions with IGNORE_EXCEPTIONS or DJANGO_ASYNC_REDIS_IGNORE_EXCEPTIONS, you may optionally log exceptions using the global variable DJANGO_ASYNC_REDIS_LOG_IGNORED_EXCEPTIONS in your settings file:

DJANGO_ASYNC_REDIS_LOG_IGNORED_EXCEPTIONS = True

If you wish to specify the logger in which the exceptions are output, simply set the global variable DJANGO_ASYNC_REDIS_LOGGER to the string name and/or path of the desired logger. This will default to __name__ if no logger is specified and DJANGO_ASYNC_REDIS_LOG_IGNORED_EXCEPTIONS is True:

DJANGO_ASYNC_REDIS_LOGGER = 'some.specified.logger'

Infinite timeout

django-async-redis comes with infinite timeouts support out of the box. And it behaves in the same way as the Django BaseCache backend specifies:

• timeout=0 expires the value immediately.
• timeout=None infinite timeout

await cache.aset("key", "value", timeout=None)

Get ttl (time-to-live) from key

With Redis, you can access to ttl of any stored key, for it, django-async-redis exposes attl function.

It returns:

• 0 if key does not exists (or already expired).
• None for keys that exists but does not have any expiration.
• ttl value for any volatile key (any key that has expiration).

>>> from django.core.cache import cache
>>> await cache.aset("foo", "value", timeout=25)
>>> await cache.attl("foo")
25
>>> await cache.attl("not-existent")
0

Expire & Persist

Additionally to the simple ttl query, you can send persist a concrete key or specify a new expiration timeout using the apersist and aexpire methods:

>>> await cache.aset("foo", "bar", timeout=22)
>>> await cache.attl("foo")
22
>>> await cache.apersist("foo")
>>> await cache.attl("foo")
None

>>> await cache.aset("foo", "bar", timeout=22)
>>> await cache.aexpire("foo", timeout=5)
>>> await cache.attl("foo")
5

Scan & Delete keys in bulk

django-async-redis comes with some additional methods that help with searching or deleting keys using glob patterns.

>>> from django.core.cache import cache
>>> await cache.akeys("foo_*")
["foo_1", "foo_2"]

A simple search like this will return all matched values. In databases with a large number of keys this isn't suitable method. Instead, you can use the aiter_keys function that works like the akeys function but uses Redis server side cursors. Calling aiter_keys will return a generator that you can then iterate over efficiently.

>>> from django.core.cache import cache
>>> await cache.aiter_keys("foo_*")
<async_generator object algo at 0x7ffa9c2713a8>
>>> (await cache.aiter_keys("foo_*")).__anext__()
"foo_1"

For deleting keys, you should use adelete_pattern which has the same glob pattern syntax as the akeys function and returns the number of deleted keys.

>>> from django.core.cache import cache
>>> await cache.adelete_pattern("foo_*")

Redis native commands

django-async-redis has limited support for some Redis atomic operations, such as the commands SETNX and INCR.

You can use the SETNX command through the backend aset() method with the nx parameter:

>>> from django.core.cache import cache
>>> await cache.aset("key", "value1", nx=True)
True
>>> await cache.aset("key", "value2", nx=True)
False
>>> await cache.aget("key")
"value1"

Also, the aincr and adecr methods use Redis atomic operations when the value that a key contains is suitable for it.

Note that setting xx to True overrides the nx flag according to aioredis.

Connection pools

Behind the scenes, django-async-redis uses the underlying aioredis connection pool implementation and exposes a simple way to configure it. Alternatively, you can directly customize a connection/connection pool creation for a backend.

The default aioredis behavior is to not close connections, recycling them when possible.

Notes

Since the majority of this code was ported from django-redis, there was one case that had needed a monkeypatch. In django_async_redis.util, we implement CacheKey which subclasses str which helps us know if a cache key was already created. Since aioredis, checks if the cache key is of type str (and others), I had to monkeypatch that check so that a CacheKey instance could also be accepted.

Credit

• Hey, I'm Andrew. I'm busy in college, but I wanted to help contribute to Django's async ecosystem.
• Lots of code and docs is taken from django-redis, including the tests. I just needed to port everything to asyncio and aioredis.
• I used cookiecutter-pypackage to generate this project.
• Thank you to Python Discord server's async topical chat for helping me understand when to use coroutines over sync functions and @Bast and @hmmmm in general because they're OG.