MLFlow UI issue when running in docker

[Darin Plutchok]

I installed mlflow and started the ui with no issues on my windows 10 machine in an anaconda python 3.7 environment and am able to access the UI via http://localhost:5000
However, when doing the exact same thing within the anacanda3 docker container the UI doesn't appear to be rendering/responding.
My docker run command includes the proper port and I am able to exec into the container, install mlflow and start the ui (>mlflow ui) without any errors.

docker run --name conda3 -d -t -v //c//develop:/develop -p 5000:5000 continuumio/anaconda3\

When I try to access http://localhost:5000 (or http://127.0.0.1:5000/), the response in the browser is "ERR_EMPTY_RESPONSE".
I tried to access the UI within the contain via lynx, to confirm that it is running, however, the response is just a warning that this site is javascript which can't be rendered in lynx.

I also tried the above docker approach using python 3.6 and also utilizing the docker container I normally develop with but in both cases I still get "ERR_EMPTY_RESPONSE".

I am pretty much stuck at this point so any suggestions will be appreciated. Thanks.

[Darin Plutchok]

Solution was to start the server with the 'server' command instead of the 'ui' command like this: mlflow server -h 0.0.0.0