Geek asks and answers

A combination of celery

I am working on a celery software update system. I have a precedent that I am struggling to implement. Here are my assignments:

device_software_updates (device_id)

returns a list of software updates that need to be installed on a device

installed_device_software (device_id)

returns the software modules that are currently installed on a device

latest_device_software (device_id)

returns the latest software versions available for a device

software_updates (installed_software, latest_software)

returns the latest software modules that are not installed

In pure python, the implementation of device_software_updates might look like

def device_software_updates(device_id):
    return software_updates(installed_device_software(device_id),
                            latest_device_software(device_id))

What is the cleanest way to implement this in Celery 3.0? I would like to do something using groups. My current implementation is as follows:

def device_software_updates(device_id):
    return (
        group(installed_device_software.s(device_id),
              latest_device_software.s(device_id)) |
        software_updates.s()
    )()

Unfortunately, this means that the software_updates argument is software_updates(arg_list)not perfect.

+5
source share
1 answer

, .

http://docs.celeryproject.org/en/latest/userguide/canvas.html#groups,

- , tasket .

...

, . , - , .

( Celery)

callback = tsum.subtask()
header = [add.subtask((i, i)) for i in xrange(100)]
result = chord(header)(callback)
result.get()

- , :

@celery.task
def device_software_updates():
    callback = software_updates.subtask()
    header = [
              installed_device_software.subtask(device_id), 
              latest_device_software.s(device_id) 
             ]
    result = chord(header)(callback)
    return result.get()
+6

More articles:


All Articles