django-dumbwaiter 0.1

Asynchronous function execution with cached results
Optimizing expensive queries and calculations in web applications is a complex and tedious problem. django-dumbwaiter is designed to enable simple caching of data while relieving the dogpile effect.

Numerous functions can be run in the background on a periodic basis, and will persist the results of their evaluation to the database.

How to use dumbwaiter

The easiest way to get started with dumbwaiter is to add `dumbwaiter` to the INSTALLED_APPS list in your `` and to add a setting called `DUMBWAITER_FUNCTION_LIST` that conforms to the following standard:


            'function': 'time.time',
            'name': 'time',
            'frequency': 5,
            'function': 'my_app.utils.function_name',
            'name': 'count',
            'frequency': 15,

Each member of the function list can contain the following attributes:

    An actual function or a string containing the path by which a function can be accessed.

    The name by which the function will be referenced. This name must be unique.

    The number of seconds between invocations of `function`. Defaults to 'DUMBWAITER_DEFAULT_FREQUENCY`.

    The arguments to be passed to `function`. Defaults to an empty list.

    The keyword arguments to be passed to `function`. Defaults to an empty dict.

    The depth of the history of cached data. This data is not accessible through the standard API, and currently only applied when using the database storage backend. Defaults to `DUMBWAITER_DEFAULT_SAVED`.

After the function list has been established the `run_dumbwaiter` management command will run the specified functions at the specified intervals.


    ./ run_dumbwaiter

Once the run_dumbwaiter management command is started, it will begin to record the values for each function. It is possible to retrieve the most recent value through the use of `get_value`:


    import dumbwaiter
    my_value = dumbwaiter.get_value("count")

Additional settings

There are several additional settings that may be customized.

    The name of a module containing the serializer to be used to store the value returned by functions in the function list. The serializer must be named `serializer` in the module specified. It must additionally support the `serialize` and `deserialize` methods. The default serializer lives in `dumbwaiter.serializers.pickler`.

    The Dumbwaiter can run in threaded and non-threaded modes. The primary advantage of threaded operation is that functions do not block one another. The default is True.

     The default depth of the history of cached data. Defaults to 10.

    The default number of seconds between invocation of functions. Defaults to five minutes.

    If using the `dumbwaiter.serializers.pickler` serializer, this specifies the version of the pickle protocol that will be used.

last updated on:
August 23rd, 2010, 13:23 GMT
license type:
BSD License 
developed by:
Ben Slavin
ROOT \ Internet \ HTTP (WWW)
Download Button

In a hurry? Add it to your Download Basket!

user rating



Rate it!

Add your review!