Skip to content

Latest commit

 

History

History
33 lines (21 loc) · 1.04 KB

README_CELERY.md

File metadata and controls

33 lines (21 loc) · 1.04 KB
Raw

How to Use it

Celery is a distributed task management framework In short, it has two components:

  • the task producer, the caller of the task, under our context, the flask app
  • the task consumer(the executor service, which is listening the task)

At the producer side, I will call it in the flask app when user is creating classifiers of running detection tasks

At the listener side, we should start a server to listen to the task

How to run the Task consumer

  • Under the director 'celery_task'
  • sudo celery -A Tasks worker --loglevel=info
  • sudo is necessary, since we are launching docker

Launching Training Job

Base: prototype image of that kind of network Task: run the training task After: after running the task, we should commit the container as an image, then remove the container

Launching Test Job

Base: image of a classifier Task: run the server, until the user remove it or the time runs up After: after it exceed, just remove the container

How to build the base image

docker build -t [name of the created image] .