Got more familiar with gunicorn by reading the docs and finding some best practices concerning tuning and optimization.
- adjusted worker count to
cpu_count() * 2 + 1(seems to be a good choice for most applications, see docs)
keep_alive(for reverse proxy)
- logs are written to the common log directory (
Other settings should be adjusted individually.
Running locally and using docker succeeded without issues and requests were logged to the according files.
Also changed makefile so that the
cp command won't exit the script (if file already exists)