django-huey-monitor

Django based tool for monitoring huey task queue

Current implementation will just store all Huey task signals into the database and display them in the Django admin.

Quickstart

pip install django-huey-monitor

INSTALLED_APPS = [
    #...
    'bx_py_utils', # https://github.com/boxine/bx_py_utils
    'huey_monitor',
    #...
]

Note: You must normally not change your Huey tasks.

Collect main-/sub-tasks

Huey monitor model can store information about task hierarchy. But this information can't be collected automatically. You have to store these information in your Task code.

e.g.:

@task(context=True)
def sub_task(task, parent_task_id, chunk_data):
    # Save the task hierarchy by:
    TaskModel.objects.set_parent_task(
        main_task_id=parent_task_id,
        sub_task_id=task.id,
    )
    # ... do something with e.g.: chunk_data ...


@task(context=True)
def main_task(task):
    for chunk_data in something:
        sub_task(parent_task_id=task.id, chunk_data=chunk_data)

Working example can be found in the test app here: huey_monitor_tests/test_app/tasks.py

Collect progress information

Running task can store progress information in a similar way as tqdm. So it's possible to see the progress in admin.

Minimal example:

@task(context=True)
def foobar_task(task, list_to_process):
    process_info = ProcessInfo(
        task,
        desc='A description of this Job',
        total=len(list_to_process)
    )

    for item in list_to_process:
        # ...to something with the {item}...
        process_info.update(n=1) # add the information that one item was processed

It is also possible to divide the work to several tasks and collect information about the processing of main-/sub-tasks.

Working example can be found in the test app here: huey_monitor_tests/test_app/tasks.py

run test project

Note: You can quickly test Huey Monitor with the test project, e.g:

~/django-huey-monitor$ ./manage.sh run_testserver

or in an isolated Docker container:

~/django-huey-monitor$ make up

More info see below.

Screenshots

(More Screenshots here: boxine.github.io/django-huey-monitor/)

2021-02-22-v030-task-details.png

2021-02-22-v030-progress-info1.png

developing

• install docker
• clone the project
• start the container

To start developing e.g.:

~$ git clone https://github.com/boxine/django-huey-monitor.git
~$ cd django-huey-monitor
~/django-huey-monitor$ make
help                 List all commands
install-poetry       install or update poetry via pip
install              install via poetry
update               Update the dependencies as according to the pyproject.toml file
lint                 Run code formatters and linter
fix-code-style       Fix code formatting
tox-listenvs         List all tox test environments
tox                  Run pytest via tox with all environments
pytest               Run pytest
pytest-ci            Run pytest with CI settings
publish              Release new version to PyPi
makemessages         Make and compile locales message files
clean                Remove created files from the test project (e.g.: SQlite, static files)
build                Update/Build docker services
up                   Start docker containers
down                 Stop all containers
shell_django         go into a interactive bash shell in Django container
shell_huey           go into a interactive bash shell in Huey worker container
logs                 Display and follow docker logs
reload_django        Reload the Django dev server
reload_huey          Reload the Huey worker
restart              Restart the containers
fire_test_tasks      Call "fire_test_tasks" manage command to create some Huey Tasks

~/django-huey-monitor$ make install-poetry
~/django-huey-monitor$ make install
~/django-huey-monitor$ make up

It's also possible to run the test setup with SQLite and Huey immediate setup without docker:

~$ git clone https://github.com/boxine/django-huey-monitor.git
~$ cd django-huey-monitor
~/django-huey-monitor$ ./manage.sh run_testserver

History

• dev
  • tbc
• v0.3.0 - 22.02.2020
  • Store and display running task progress information a little bit as tqdm #17
• v0.2.0 - 17.02.2020
  • Store "parent_task" information for main-/sub-tasks
• v0.1.0 - 21.12.2020
  • Work-a-round if a Huey worker died
  • Fix missing translations from bx_py_utils in test project
  • Simulate a cluster of Huey worker via docker and the test project
• v0.0.1 - 15.12.2021
  • initial release

License

GPL. Patches welcome!

Links

• https://pypi.org/project/django-huey-monitor/