Django KCK
Django KCK is data orchestration for Django. It can be used for:
- scheduled data imports from remote sources
- ensuring each data product kept fresh, either by updating at a regular interval or when there is a change in source data on upon which it depends
- preparing complex data products in advance of a likely request
- simplifying and optimizing complex data flows
The development pattern Django KCK encourages for data products emphasizes compartmentalization and simplification over complexity, cached data with configurable refresh routines over real-time computation, and common-sense optimizations over sprawling distributed parallelism.
History
Django KCK is a simplified version of KCK that targets the Django environment exclusively. It also uses PostgreSQL as the cache backend, instead of Cassandra.
Quick Install
Basic Usage
# myapp/primers.py
from kck import Primer
class TitleListPrimer(Primer):
key = 'title_list'
parameters = [
{"name": "id", "from_str": int}
]
def compute(self, key):
param_dict = self.key_to_param_dict(key)
results = [{ 'title': lkp_title(id) } for id in param_dict['id_list']]
return results
# myapp/views.py
from kck import Cache
from django.http import JsonResponse
def first_data_product_view(request, author_id):
cache = Cache.get_instance()
title_list = cache.get(f'title_list/{author_id}')
return JsonResponse(title_list)
Theory
Essentially, Django KCK is a lazy-loading cache. Instead of warming the cache in advance, Django KCK lets a developer tell the cache how to prime itself in the event of a cache miss.
If we don't warm the cache in advance and we ask the cache for a data product that depends on a hundred other data products in the cache, each of which either gathers or computes data from other sources, then this design will only generate or request the data that is absolutely necessary for the computation. In this way, Django KCK is able to do the last amount of work possible to accomplish the task.
To further expedite the process or building derivative data products, Django KCK includes mechanisms that allow for periodic or triggered updates of data upon which a data product depends, such that it will be immediately available when a request is made.
It also makes it possible to "augment" derivative data products with new information so that, for workloads that can take advantage of the optimization, a data product can be updated in place, without regenerating the product in its entirety. Where it works, this approach can turn minutes of computation into milliseconds.
1 maintainer
Metadata
1 maintainer
Release files for django-kck 0.0.56
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| django-kck-0.0.56.tar.gz | 12.5 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| django_kck-0.0.56-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 28.9 kB
Release files / django-kck-0.0.56.tar.gz
| Download URL | django-kck-0.0.56.tar.gz |
|---|---|
| Size | 12.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
a48cb5b133a6cdf0454c7736a6306cf2bccbb8dabb0917257847a832c0b02eaa
|
|
BLAKE2b-256 checksum How to use checksums |
25c49b22b22d197fe6fd4fa2936fdc9820559c2a578a2d4ace90187930ac59f7
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.36.1 CPython/3.6.8
|
Release files / django_kck-0.0.56-py3-none-any.whl
| Download URL | django_kck-0.0.56-py3-none-any.whl |
|---|---|
| Size | 16.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
8b1b403bfac6967c34da5ef8d708b44d9b0ac274f43656dcfb57d3618be687e1
|
|
BLAKE2b-256 checksum How to use checksums |
8d8d52f32b39f40e7dc9acc7fea8033797901c8b9618d48aa3b18fab89ac76ad
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.36.1 CPython/3.6.8
|
