We Keep Coding

Azure App Service: specifying docker run port on container side

I want to use scrapinghub/splash container on Azure App Service (Web App for Containers) on Linux.
But docker run command on deploy randomly changes the binding port of container side (see the log below, port 8961 is automatically assigned. this number varies every deploy)
2020-01-21 08:56:47.494 INFO - docker run -d -p 8961:8050 --name b2scraper-splash_3_d89ce1f2 -e WEBSITES_ENABLE_APP_SERVICE_STORAGE=false -e WEBSITES_PORT=8050 -e WEBSITE_SITE_NAME=b2scraper-splash -e WEBSITE_AUTH_ENABLED=False -e PORT=8050 -e WEBSITE_ROLE_INSTANCE_ID=0 -e WEBSITE_HOSTNAME=b2scraper-splash.azurewebsites.net -e WEBSITE_INSTANCE_ID=5446f93a2cbcb25300f091395c54ce738773ce47489c2818322ffabbc23e3413 scrapinghub/splash:latest python3 /app/bin/splash --proxy-profiles-path /etc/splash/proxy-profiles --js-profiles-path /etc/splash/js-profiles --filters-path /etc/splash/filters --lua-package-path "/etc/splash/lua_modules/?.lua" --disable-private-mode --port 8050
Changing host port binding is possible using WEBSITES_PORT, but seems no way to change container side.
Is there way to fix container-side port binding like -p 8050:80 or -p 8050:443 of docker run command?
e.g. Using the container on Azure Container Instances is possible, without changing service port 8050.

--publish in the docker run command creates a firewall rule which maps a container port to a port on the Docker host.
https://docs.docker.com/config/containers/container-networking/
For the command: docker run -d -p 8961:8050 imagename, TCP port 8050 in the container is mapped to 8961 on the Docker Host. On App Services, this docker run command cannot be changed. The container port (8050 in this case) can be set to a specific value using WEBSITES_PORT application setting.

That doesn´t work. you get 443 as port with HTTPS.
Neither EXPOSE XXXX, nor WEBSITES_PORT or PORT as configuration parameters...
You do see "docker run -d -p 8961:8050" in the logs, but it doesn´t matter to Azure when it comes to exposing the app...

docker port exposing issue

I have a flask application that I'm trying to dockerize but the ports are not getting exposed properly.
DockerFile
FROM tiangolo/uwsgi-nginx-flask:python3.7
LABEL Name=testAPP Version=0.0.1
EXPOSE 5000
ADD . /app
WORKDIR /app
# Using pip:
RUN python3 -m pip install -r requirements.txt
ENTRYPOINT [ "python3" ]
CMD ["application.py" ,"runserver","-h 0.0.0.0"]
Docker Build is successful:
docker build --rm -f "Dockerfile" -t testAPP .
Docker Run is building the image successfully
docker run -device -expose 5000:5000 testAPP
Also tried,
docker run --rm -d -p 443:443/tcp -p 5000:5000/tcp -p 80:80/tcp testAPP
But when I try to access the site it throws an error
site can't be reached error
Flask App(Inside the APP)
if __name__ == '__main__':
app.run(host='127.0.0.1', port=5000)
On Execution of the command
Docker container ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
8724cdb38e14 testAPP "/entrypoint.sh pyth…" 15 seconds ago Up 13 seconds 80/tcp, 443/tcp, 0.0.0.0:5000->5000/tcp funny_galois

Defining a port as exposed doesn’t publish the port by itself. Try with the flag -p , like:
-p container_port:local_port
example:
docker run -p 8080:8080 -v ~/Code/PYTHON/ttftt-recipes-manager:/app python_dev
But before running try to check if there is something else that already running on the specified port like:
lsof -i :PORTNUM
and after with something like:
docker logs my_container

Make sure you're mapping your localhost port to the container's port
docker run -p 127.0.0.1:8000:8000 your_image
And once you're application is in the container, you want to run your app with the host set to 0.0.0.0

Run Python from Docker

I'm trying out Docker these days and I want to run create virtual environments in Python in Docker. I downloaded Miniconda3 from docker hub and tested out with basic hello world program written in python.
I ran:
docker run -i-t continuumio/miniconda3 /bin/bash
Then on another terminal I ran:
docker exec laughing_wing "python ~/Documents/Test/hello_world.py"
Where the name of docker container is laughing_wing, and my hello_world.py is in Documents/Test directory.
But running the second command I get:
"OCI runtime exec failed: exec failed: container_linux.go:344:
starting container process caused "exec: \"python
~/Documents/Test/hello_world.py\": stat python
~/Documents/Test/hello_world.py: no such file or directory": unknown"
I'm confused about this.

Looks like you're trying to have the docker container run a python file from your machine. The docker container is isolated from it's host, so you need to either create your own Docker image where you add the file, or mount the ~/Documents/Test directory to your docker container. Something like this:
docker run -it -v ~/Documents/Test:/Test continuumio/miniconda3 /bin/bash
docker exec *container_name* "python /Test/hello_world.py"

Running Jupyter notebook in Docker

I want to run jupyter in docker container. I am not able to launch the jupyter notebook. When I copy paste the URL given in the terminal.. server cannot be reached. Will appreciate any ideas to try

You are forwarding port 8080 in the docker run call with -p 8080:8080. But you also need to forward port 8888 by adding -p 8888:8888. More specifically, you want to run:
docker -it -p 8080:8080 -p 8888:8888 jupyter/minimal-notebook

First thing is Jupyter nootbook runs on port 8888. If you want to access the notebook on a diff port on your host you should map it like this -p 80:8888.
If you don't mind using the defaults, you should use run the following command.
Run this command: docker run -p 8888:8888 jupyter/minimal-notebook. Then
replace the host name in the url given in terminal with localhost like this
http://localhost:8888/\?token\=<TOKEN>\&token\=<TOKEN>
This should work.
Note: If you map it to a diff port, you should replace it in the url you get in the terminal. Ex. http://localhost:80/\?token\=<TOKEN>\&token\=<TOKEN>

Docker with Jupyter Notebook and Spark

I am trying to use docker to run Jupyter Notebook with spark , on PowerShell I am using the following command:
docker run --rm -h 0.0.0.0 -p 8888:8888 -e PYTHONHASHSEED=0 jupyter/all-spark-notebook
When I run that I get the following error:
Container must be run with group "root" to update passwd file
Any Ideas about how to fix it/ update group root password?