CODEWITHSUNDEEP
multithreadingherokugunicorn
Jonathan Livni
Jonathan Livni

Reputation: 107232

gunicorn threads on Heroku

I'm trying to deploy a python web application on Heroku using the following Procfile command:

web: gunicorn service:app --log-file=- --workers 1 --threads 4

This generates the following error:

Traceback (most recent call last):
  File "/app/.heroku/python/lib/python2.7/site-packages/gunicorn/util.py", line 139, in load_class
    mod = import_module('.'.join(components))
  File "/app/.heroku/python/lib/python2.7/importlib/__init__.py", line 37, in import_module
    __import__(name)
  File "/app/.heroku/python/lib/python2.7/site-packages/gunicorn/workers/gthread.py", line 37, in <module>
    """)
RuntimeError:
    You need 'concurrent' installed to use this worker with this python version.

As we do not control the gunicorn installation on Heroku, how do I resolve this dependency issue?

Oddly, Heroku do not mention gunicorn threads in their docs, only workers. Is this related?

Upvotes: 5

Views: 2324

Answers (1)

mehdix
mehdix

Reputation: 5164

Add these lines to your requirements.txt:

futures==2.1.6
trollius==1.0.1

After pushing these changes to your heroku instance you will be able to run gunicorn with threads.

concurrent is a python 3.2 package and is backported to python 2.7 as future package. You should either use at least python 3.2 as your runtime or you should add the backported packages to your project requirements.

Upvotes: 9

Related Questions