Django, ImportError: cannot import name Celery, possible circular import?
Question:
I went through this example here:
http://docs.celeryproject.org/en/latest/django/first-steps-with-django.html
All my tasks are in files called tasks.py.
After updating celery and adding the file from the example django is throwing the following error, no matter what I try:
ImportError: cannot import name Celery
Is the problem possibly caused by the following?
app.autodiscover_tasks(settings.INSTALLED_APPS, related_name='tasks')
Because it goes through all tasks.py files which all have the following import.
from cloud.celery import app
cloud/celery.py:
from __future__ import absolute_import
import os, sys
from celery import Celery
from celery.schedules import crontab
from django.conf import settings
BROKER_URL = 'redis://:PASSWORD@localhost'
os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'cloud.settings')
app = Celery('cloud', broker=BROKER_URL)
app.config_from_object('django.conf:settings')
app.autodiscover_tasks(settings.INSTALLED_APPS, related_name='tasks')
if "test" in sys.argv:
app.conf.update(
CELERY_ALWAYS_EAGER=True,
)
print >> sys.stderr, 'CELERY_ALWAYS_EAGER = True'
CELERYBEAT_SCHEDULE = {
'test_rabbit_running': {
"task": "retail.tasks.test_rabbit_running",
"schedule": 3600, #every hour
},
[..]
app.conf.update(
CELERYBEAT_SCHEDULE=CELERYBEAT_SCHEDULE
)
retail/tasks.py:
from cloud.celery import app
import logging
from celery.utils.log import get_task_logger
logger = get_task_logger('tasks')
logger.setLevel(logging.DEBUG)
@app.task
def test_rabbit_running():
import datetime
utcnow = datetime.datetime.now()
logger.info('CELERY RUNNING')
The error happens, when I try to access a url that is not valid, like /foobar.
Here is the full traceback:
Traceback (most recent call last):
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/gunicorn/workers/sync.py", line 126, in handle_request
respiter = self.wsgi(environ, resp.start_response)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/wsgi.py", line 255, in __call__
response = self.get_response(request)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/base.py", line 178, in get_response
response = self.handle_uncaught_exception(request, resolver, sys.exc_info())
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/base.py", line 220, in handle_uncaught_exception
if resolver.urlconf_module is None:
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/urlresolvers.py", line 342, in urlconf_module
self._urlconf_module = import_module(self.urlconf_name)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/utils/importlib.py", line 35, in import_module
__import__(name)
File "/opt/src/slicephone/cloud/cloud/urls.py", line 52, in
urlpatterns += patterns('', url(r'^search/', include('search.urls')))
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/conf/urls/__init__.py", line 25, in include
urlconf_module = import_module(urlconf_module)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/utils/importlib.py", line 35, in import_module
__import__(name)
File "/opt/src/slicephone/cloud/search/urls.py", line 5, in
from handlers import SearchHandler
File "/opt/src/slicephone/cloud/search/handlers.py", line 15, in
from places import handlers as placeshandler
File "/opt/src/slicephone/cloud/places/handlers.py", line 23, in
import api as placesapi
File "/opt/src/slicephone/cloud/places/api.py", line 9, in
from djapi import *
File "/opt/src/slicephone/cloud/places/djapi.py", line 26, in
from tasks import add_single_place, add_multiple_places
File "/opt/src/slicephone/cloud/places/tasks.py", line 2, in
from cloud.celery import app
File "/opt/src/slicephone/cloud/cloud/celery.py", line 4, in
from celery import Celery
File "/opt/src/slicephone/cloud/cloud/celery.py", line 4, in
from celery import Celery
ImportError: cannot import name Celery
Answers:
Did you add the line:
from __future__ import absolute_import
to the top of your cloud/celery.py module?
Read the breakdown of the example here:
http://docs.celeryproject.org/en/latest/django/first-steps-with-django.html
Adding the following lines to cloud/celery.py:
import celery
print celery.__file__
gave me the file itself and not the celery module from the library. After renaming celery.py to celeryapp.py and adjusting the imports all errors were gone.
Note:
That leads to a change in starting the worker:
celery worker --app=cloud.celeryapp:app
For those running celery==3.1.2 and getting this error:
TypeError: unpack_from() argument 1 must be string or read-only buffer, not memoryview
Apply the patch mentioned here: https://github.com/celery/celery/issues/1637
I got the same error.
Seems that from __future__ import absolute_import DOES NOT work for Python 2.6.1, still not raising an error.
Upgraded to Python 2.7.5 and it just worked.
For someone who want to know what cause this error:
I have meet this problem just now, then I found the problem — sys.path.
Maybe you add some path to sys.path like me, I add below code in manage.py,
ROOT_PATH = os.path.dirname(os.path.abspath(__file__))
SRC_PATH = os.path.join(ROOT_PATH, 'src')
CONF_PATH = os.path.join(ROOT_PATH, 'conf')
sys.path.insert(0, SRC_PATH)
sys.path.insert(0, CONF_PATH)
so, from celery import Celery would search celery in SRC_PATH and CONF_PATH first, that’s the problem.
change to
sys.path.append(SRC_PATH)
sys.path.append(CONF_PATH)
It would search in python’s lib and site-packages first. Solved perfectly.
Note that older Django projects have the manage.py script in the same directory as the project directory. That is, the structure looks like this:
- proj/
- proj/__init__.py
- proj/celery.py
- proj/urls.py
- proj/manage.py
- proj/settings.py
instead of this:
- proj/
- proj/__init__.py
- proj/celery.py
- proj/settings.py
- proj/urls.py
- manage.py
In this case, you will just have to rename the celery.app file to something different, like celeryapp.py as suggested in the accepted answer above.
With Django 1.7.5, Celery 3.1.17, and Python 2.7.6 I found that I was still getting these ImportError: cannot import name Celery. But only when running tests under PyCharm 4.0.4.
I found that a solution was not to rely on from __future__ import absolute_import as described in First Steps with Django. Instead I renamed proj/proj/celery.py to proj/proj/celery_tasks.py and then changed the content of __init__.py to match: from .celery_tasks import app as celery_app. No more multiple instances of files named celery.py to cause import confusion seemed to be a simpler approach.
I got the same error. It turns out there was a problem with my Celery version. I upgraded to 3.1 and celeryd is now deprecated for this version (http://celery.readthedocs.org/en/latest/whatsnew-3.1.html). So I had to downgrade to the version 3.0.19 that was the previous stable version used for the project, and it works good so far.
pip install celery==3.0.19
Anyway, if you don’t want to downgrade, the replacement for celeryd in the version 3.1 is celery worker. Check here for more info: http://celery.readthedocs.org/en/latest/userguide/workers.html.
Hope this helps! 🙂
Work for me ( some bug after deploy in server ):
Remove all *.pyc files from project and restart him.
got the same error
my celery settings filename which was(celery.py) was conflicting with ‘celery’ package…
so while doing this-> from celery import Celery ,
it was raising error- cannot import name Celery
solution->just change the ‘celery.py’ to something else like ‘celery-settings.py’
I have face similar type of issue:
from celery import Celery
ImportError: cannot import name ‘Celery’ from ‘celery’
Another simple way to solve this:
If your package have celery configuration in celery.py this is the reason that it is causing problems.
Rename it something like celery_settings.py
In my django application I had the same error. Version Django 4, Celery 5. My my problem is that I run Celery from directory with file celery.py, but it is necessary from directory with manage.py.
Slava Ukraini!
In October 2022, importlib-metadata==5.0.0 was released. In python 3.7, this breaks kombu==5.2.4 (see issue#1600) which is used by the current version of celery.
Pin importlib-metadata==4.13.0 or another version less than 5, or update to python 3.8.
I faced the same issue on a FastAPI app running on Python 3.7. None of the solutions from this thread were working for me. I fixed the issue on my app by upgrading Python to 3.8.
If the above error i.e. ImportError: cannot import name 'Celery' from 'celery' (/usr/local/airflow/.local/lib/python3.7/site-packages/celery/__init__.py) is coming and the python version is being used as 3.7.*, then due to a bug in importlib-metadata library because of its latest release.
So, to fix the above issue please use this version in your python dependencies list:
- pip install importlib-metadata==4.13.0
Alternately, if you are using Docker file to build the project, change the python version from 3.7 to 3.8 and it should work fine.
For me, it worked. Please use the below link for a detailed look:
I went through this example here:
http://docs.celeryproject.org/en/latest/django/first-steps-with-django.html
All my tasks are in files called tasks.py.
After updating celery and adding the file from the example django is throwing the following error, no matter what I try:
ImportError: cannot import name Celery
Is the problem possibly caused by the following?
app.autodiscover_tasks(settings.INSTALLED_APPS, related_name='tasks')
Because it goes through all tasks.py files which all have the following import.
from cloud.celery import app
cloud/celery.py:
from __future__ import absolute_import
import os, sys
from celery import Celery
from celery.schedules import crontab
from django.conf import settings
BROKER_URL = 'redis://:PASSWORD@localhost'
os.environ.setdefault('DJANGO_SETTINGS_MODULE', 'cloud.settings')
app = Celery('cloud', broker=BROKER_URL)
app.config_from_object('django.conf:settings')
app.autodiscover_tasks(settings.INSTALLED_APPS, related_name='tasks')
if "test" in sys.argv:
app.conf.update(
CELERY_ALWAYS_EAGER=True,
)
print >> sys.stderr, 'CELERY_ALWAYS_EAGER = True'
CELERYBEAT_SCHEDULE = {
'test_rabbit_running': {
"task": "retail.tasks.test_rabbit_running",
"schedule": 3600, #every hour
},
[..]
app.conf.update(
CELERYBEAT_SCHEDULE=CELERYBEAT_SCHEDULE
)
retail/tasks.py:
from cloud.celery import app
import logging
from celery.utils.log import get_task_logger
logger = get_task_logger('tasks')
logger.setLevel(logging.DEBUG)
@app.task
def test_rabbit_running():
import datetime
utcnow = datetime.datetime.now()
logger.info('CELERY RUNNING')
The error happens, when I try to access a url that is not valid, like /foobar.
Here is the full traceback:
Traceback (most recent call last):
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/gunicorn/workers/sync.py", line 126, in handle_request
respiter = self.wsgi(environ, resp.start_response)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/wsgi.py", line 255, in __call__
response = self.get_response(request)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/base.py", line 178, in get_response
response = self.handle_uncaught_exception(request, resolver, sys.exc_info())
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/handlers/base.py", line 220, in handle_uncaught_exception
if resolver.urlconf_module is None:
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/core/urlresolvers.py", line 342, in urlconf_module
self._urlconf_module = import_module(self.urlconf_name)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/utils/importlib.py", line 35, in import_module
__import__(name)
File "/opt/src/slicephone/cloud/cloud/urls.py", line 52, in
urlpatterns += patterns('', url(r'^search/', include('search.urls')))
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/conf/urls/__init__.py", line 25, in include
urlconf_module = import_module(urlconf_module)
File "/opt/virtenvs/django_slice/local/lib/python2.7/site-packages/django/utils/importlib.py", line 35, in import_module
__import__(name)
File "/opt/src/slicephone/cloud/search/urls.py", line 5, in
from handlers import SearchHandler
File "/opt/src/slicephone/cloud/search/handlers.py", line 15, in
from places import handlers as placeshandler
File "/opt/src/slicephone/cloud/places/handlers.py", line 23, in
import api as placesapi
File "/opt/src/slicephone/cloud/places/api.py", line 9, in
from djapi import *
File "/opt/src/slicephone/cloud/places/djapi.py", line 26, in
from tasks import add_single_place, add_multiple_places
File "/opt/src/slicephone/cloud/places/tasks.py", line 2, in
from cloud.celery import app
File "/opt/src/slicephone/cloud/cloud/celery.py", line 4, in
from celery import Celery
File "/opt/src/slicephone/cloud/cloud/celery.py", line 4, in
from celery import Celery
ImportError: cannot import name Celery
Did you add the line:
from __future__ import absolute_import
to the top of your cloud/celery.py module?
Read the breakdown of the example here:
http://docs.celeryproject.org/en/latest/django/first-steps-with-django.html
Adding the following lines to cloud/celery.py:
import celery
print celery.__file__
gave me the file itself and not the celery module from the library. After renaming celery.py to celeryapp.py and adjusting the imports all errors were gone.
Note:
That leads to a change in starting the worker:
celery worker --app=cloud.celeryapp:app
For those running celery==3.1.2 and getting this error:
TypeError: unpack_from() argument 1 must be string or read-only buffer, not memoryview
Apply the patch mentioned here: https://github.com/celery/celery/issues/1637
I got the same error.
Seems that from __future__ import absolute_import DOES NOT work for Python 2.6.1, still not raising an error.
Upgraded to Python 2.7.5 and it just worked.
For someone who want to know what cause this error:
I have meet this problem just now, then I found the problem — sys.path.
Maybe you add some path to sys.path like me, I add below code in manage.py,
ROOT_PATH = os.path.dirname(os.path.abspath(__file__))
SRC_PATH = os.path.join(ROOT_PATH, 'src')
CONF_PATH = os.path.join(ROOT_PATH, 'conf')
sys.path.insert(0, SRC_PATH)
sys.path.insert(0, CONF_PATH)
so, from celery import Celery would search celery in SRC_PATH and CONF_PATH first, that’s the problem.
change to
sys.path.append(SRC_PATH)
sys.path.append(CONF_PATH)
It would search in python’s lib and site-packages first. Solved perfectly.
Note that older Django projects have the manage.py script in the same directory as the project directory. That is, the structure looks like this:
- proj/
- proj/__init__.py
- proj/celery.py
- proj/urls.py
- proj/manage.py
- proj/settings.py
instead of this:
- proj/
- proj/__init__.py
- proj/celery.py
- proj/settings.py
- proj/urls.py
- manage.py
In this case, you will just have to rename the celery.app file to something different, like celeryapp.py as suggested in the accepted answer above.
With Django 1.7.5, Celery 3.1.17, and Python 2.7.6 I found that I was still getting these ImportError: cannot import name Celery. But only when running tests under PyCharm 4.0.4.
I found that a solution was not to rely on from __future__ import absolute_import as described in First Steps with Django. Instead I renamed proj/proj/celery.py to proj/proj/celery_tasks.py and then changed the content of __init__.py to match: from .celery_tasks import app as celery_app. No more multiple instances of files named celery.py to cause import confusion seemed to be a simpler approach.
I got the same error. It turns out there was a problem with my Celery version. I upgraded to 3.1 and celeryd is now deprecated for this version (http://celery.readthedocs.org/en/latest/whatsnew-3.1.html). So I had to downgrade to the version 3.0.19 that was the previous stable version used for the project, and it works good so far.
pip install celery==3.0.19
Anyway, if you don’t want to downgrade, the replacement for celeryd in the version 3.1 is celery worker. Check here for more info: http://celery.readthedocs.org/en/latest/userguide/workers.html.
Hope this helps! 🙂
Work for me ( some bug after deploy in server ):
Remove all *.pyc files from project and restart him.
got the same error
my celery settings filename which was(celery.py) was conflicting with ‘celery’ package…
so while doing this-> from celery import Celery ,
it was raising error- cannot import name Celery
solution->just change the ‘celery.py’ to something else like ‘celery-settings.py’
I have face similar type of issue:
from celery import Celery
ImportError: cannot import name ‘Celery’ from ‘celery’
Another simple way to solve this:
If your package have celery configuration in celery.py this is the reason that it is causing problems.
Rename it something like celery_settings.py
In my django application I had the same error. Version Django 4, Celery 5. My my problem is that I run Celery from directory with file celery.py, but it is necessary from directory with manage.py.
Slava Ukraini!
In October 2022, importlib-metadata==5.0.0 was released. In python 3.7, this breaks kombu==5.2.4 (see issue#1600) which is used by the current version of celery.
Pin importlib-metadata==4.13.0 or another version less than 5, or update to python 3.8.
I faced the same issue on a FastAPI app running on Python 3.7. None of the solutions from this thread were working for me. I fixed the issue on my app by upgrading Python to 3.8.
If the above error i.e. ImportError: cannot import name 'Celery' from 'celery' (/usr/local/airflow/.local/lib/python3.7/site-packages/celery/__init__.py) is coming and the python version is being used as 3.7.*, then due to a bug in importlib-metadata library because of its latest release.
So, to fix the above issue please use this version in your python dependencies list:
- pip install importlib-metadata==4.13.0
Alternately, if you are using Docker file to build the project, change the python version from 3.7 to 3.8 and it should work fine.
For me, it worked. Please use the below link for a detailed look: