This work on unscaled applications but only as a plugin cartridge on scaled applications. This means that it's can't scale independently of your main gear. At this point having a celery cartridge is no better than just starting celery in your app.y or action hooks. Further development can combine this with a python cartridge so it can operate in it's own gears.

Instead of using this I recommend starting celery in your action hooks or as part of you app.py instead. This will give more flexibility, and example is here https://github.com/appsembler/appsembler-launch-openshift

Configuration

OPENSHIFT_CELERY_CONFIG This is the name of your config file, which by default is $OPENSHIFT_CELERY_DIR/conf.d/celeryconfig.py but you can copy this to $OPENSHIFT_DATA_DIR/config/celeryconfig.py and edit your own copy.

To install

rhc cartridge-add https://raw.github.com/wassname/openshift-celery-cartridge/master/metadata/manifest.yml -a "appname"

Any log output will be generated to ${OPENSHIFT_HOMEDIR}logs/celery_log.txt and will be viewable with the rhc tail "appname" command

This was tested using:

celery==3.1.11
redis==2.10.3

To manage

$ rhc cartridge-status celeryd -a "yourapp"