Use async
from your code to quickly offload tasks to the Cluster
:
from django_q.tasks import async, result
# create the task
async('math.copysign', 2, -2)
# or with import and storing the id
import math.copysign
task_id = async(copysign, 2, -2)
# get the result
task_result = result(task_id)
# result returns None if the task has not been executed yet
# you can wait for it
task_result = result(task_id, 200)
# but in most cases you will want to use a hook:
async('math.modf', 2.5, hook='hooks.print_result')
# hooks.py
def print_result(task):
print(task.result)
async
can take the following optional keyword arguments:
The function to call after the task has been executed. This function gets passed the complete Task
object as its argument.
A group label. Check groups
for group functions.
Overrides the result backend's save setting for this task.
Overrides the cluster's timeout setting for this task.
Simulates a task execution synchronously. Useful for testing. Can also be forced globally via the sync
configuration option.
A broker instance, in case you want to control your own connections.
None of the option keywords get passed on to the task function. As an alternative you can also put them in a single keyword dict named q_options
. This enables you to use these keywords for your function call:
# Async options in a dict
opts = {'hook': 'hooks.print_result',
'group': 'math',
'timeout': 30}
async('math.modf', 2.5, q_options=opts)
Please not that this will override any other option keywords.
Note
For tasks to be processed you will need to have a worker cluster running in the background using python manage.py qcluster
or you need to configure Django Q to run in synchronous mode for testing using the sync
option.
You can group together results by passing async
the optional group
keyword:
# result group example
from django_q.tasks import async, result_group
for i in range(4):
async('math.modf', i, group='modf')
# after the tasks have finished you can get the group results
result = result_group('modf')
print(result)
[(0.0, 0.0), (0.0, 1.0), (0.0, 2.0), (0.0, 3.0)]
Take care to not limit your results database too much and call delete_group
before each run, unless you want your results to keep adding up. Instead of result_group
you can also use fetch_group
to return a queryset of Task
objects.:
# fetch group example
from django_q.tasks import fetch_group, count_group, result_group
# count the number of failures
failure_count = count_group('modf', failures=True)
# only use the successes
results = fetch_group('modf')
if failure_count:
results = results.exclude(success=False)
results = [task.result for task in successes]
# this is the same as
results = fetch_group('modf', failures=False)
results = [task.result for task in successes]
# and the same as
results = result_group('modf') # filters failures by default
Getting results by using result_group
is of course much faster than using fetch_group
, but it doesn't offer the benefits of Django's queryset functions.
Note
Calling Queryset.values
for the result on Django 1.7 or lower will return a list of encoded results. If you can't upgrade to Django 1.8, use list comprehension or an iterator to return decoded results.
You can also access group functions from a task result instance:
from django_q.tasks import fetch
task = fetch('winter-speaker-alpha-ceiling')
if task.group_count() > 100:
print(task.group_result())
task.group_delete()
print('Deleted group {}'.format(task.group))
async
can be instructed to execute a task immediately by setting the optional keyword sync=True
. The task will then be injected straight into a worker and the result saved by a monitor instance:
from django_q.tasks import async, fetch
# create a synchronous task
task_id = async('my.buggy.code', sync=True)
# the task will then be available immediately
task = fetch(task_id)
# and can be examined
if not task.success:
print('An error occurred: {}'.format(task.result))
An error occurred: ImportError("No module named 'my'",)
Note that async
will block until the task is executed and saved. This feature bypasses the Redis server and is intended for debugging and development. Instead of setting sync
on each individual async
you can also configure sync
as a global override.
Django Q tries to pass broker instances around its parts as much as possible to save you from running out of connections. When you are making individual calls to async
a lot though, it can help to set up a broker to reuse for async
:
# broker connection economy example
from django_q.tasks import async
from django_q.brokers import get_broker
broker = get_broker()
for i in range(50):
async('math.modf', 2.5, broker=broker)
Tip
If you are using django-redis , you can configure <django_redis>
Django Q to use its connection pool.
- param object func
The task function to execute
- param tuple args
The arguments for the task function
- param object hook
Optional function to call after execution
- param str group
An optional group identifier
- param int timeout
Overrides global cluster
timeout
.- param bool save
Overrides global save setting for this task.
- param bool sync
If set to True, async will simulate a task execution
- param redis
Optional redis connection
- param dict q_options
Options dict, overrides option keywords
- param dict kwargs
Keyword arguments for the task function
- returns
The uuid of the task
- rtype
str