Trying to run Cloudera Image in Docker

Question

I am trying to run cloudera/clusterdock in a docker image for a university project. This is my first time using docker and so far I have been using the instructions on the cloudera website which are a little sparse.

I successfully downloaded docker and the cloudera image and when I run the docker-images command I get the following:

REPOSITORY             TAG                 IMAGE ID            CREATED             SIZE
cloudera/clusterdock   latest              9b4d4f1dda22        7 days ago          467.5 MB

When I try and run up the container with this image. Using the following command

docker run cloudera/clusterdock:latest /bin/bash

I get the following message

    File "/bin/bash", line 1
    SyntaxError: Non-ASCII character '\x80' in file /bin/bash on line 2,
 but no encoding declared; see http://www.python.org/peps/pep-0263.html for details

Having read the pep mentioned pep I know I need to change the encoding in a file but the pep concentrates on python files and I am unaware of having a python file so have no idea where to find it to correct it. Also, having limited knowledge I am uneasy changing the bin/bash file as I know it can affect your machine.

Any help will have to assume I have little knowledge of this as I have little experience.

Answer 1

If you look at Dockerfile for cloudera/clusterdock:latest , you can see:

ENTRYPOINT ["python"]

So, when you do docker run cloudera/clusterdock:latest /bin/bash , you are basically doing python /bin/bash inside the container. You will see the same error if you type that in your terminal, normally:

$ python /bin/bash
  File "/bin/bash", line 1
SyntaxError: Non-ASCII character '\xe0' in file /bin/bash on line 2, but no encoding declared; see http://python.org/dev/peps/pep-0263/ for details

You probably wanted to do:

docker run -it --entrypoint=/bin/bash cloudera/clusterdock:latest

Look at clusterdock.sh to see how actually the container is supposed to be run.

Answer 2

The associated docs (eg the description on the image's Docker Hub page or our blog post) describe that clusterdock is intended to be run by sourcing clusterdock.sh. This is required because the framework controls Docker on the host machine.