Code Sample

Before going onto the details, let’s take a quick look at the Garcon’s implementation of Serial Activity Execution:

from garcon import activity
from garcon import runners


domain = 'dev'
name = 'boto_tutorial'
create = activity.create(domain, name)

a_tasks = create(
    name='a_tasks',
    run=runner.Sync(
        lambda context, activity: dict(result='Now don’t be givin him sambuca!'))

b_tasks = create(
    name='b_tasks',
    requires=[a_tasks],
    run=runner.Sync(
        lambda context, activity: print(context)))

c_tasks = create(
    name='c_tasks',
    requires=[b_tasks],
    run=runner.Sync(
        lambda context, activity: print(context)))

By way of comparison, check out the implementation using directly boto.

Note:

Notes: Executing this code shows that the activity “a_tasks” returns a dictionary which hydrates the execution context. When the activity “b_tasks” is executed, the context passed for its execution contains the key/value previously passed as an output. Same observation can be done in “c_tasks”.

All activities are running in series. More examples (including runners) are available online.