Closed alison-f-cooper closed 7 years ago
Thank you for the report. Could try to run Gunicorn with strace
? Errno 14 usually means either some invalid value was passed into a syscall or in some cases there was no memory available on the host.
Also, what's the full Python 3.5 and Gunicorn version? Is there any difference between your development machine and production server? Did you get the same error without using Docker? Could you try to run a simpler app on your production server (with using Docker) You can use https://github.com/benoitc/gunicorn/blob/master/examples/test.py as a test app.
Hi there -- I upgraded our gunicorn version to 19.7, and got some more precise output with debug logging. It looks like the problems are related to this:
https://github.com/benoitc/gunicorn/issues/1194
I will be continuing the conversation on that thread, since it seems a lot of folks are running into similar issues.
Thanks.
I'm able to run my Flask app locally using Docker, with the following command in my Dockerfile:
CMD ["gunicorn", "-w", "20", "-b", "0.0.0.0:5000", "--capture-output", "--enable-stdio-inheritance", "--access-logfile", "-", "app:app"]
However, upon attempting to deploy the application to a test server, there is an exception in the worker process, and it crashes:
Any idea what's going on here? It looks like it's trying to import the app, is hitting the app's init file, and for some reason is bombing out on some sql alchemy code in the imports section. We import Alchy in our init file, and it's trying to pull in the sql alchemy engine, and just blows up with "Bad Address."
Thanks.