Skip to content

Latest commit

 

History

History
159 lines (136 loc) · 7.95 KB

Docker2.md

File metadata and controls

159 lines (136 loc) · 7.95 KB

Your first Python container

After following the link, you will find many different versions of Python, such as 3.12, 3.9, etc. Which one should you choose? Additionally, terms like 'Alpine,' 'Bookworm,' and 'Bullseye' refer to different versions of Linux. Alpine is considered extremely lightweight, while Bookworm and Bullseye are versions of Debian (cousin of Ubuntu)." I randomly select "bookworm" python 3.9.19 and click on the link, which brings me to the Dockerfile.

Dockerfile

A Dockerfile is a text file that contains instructions on how Docker should build a container. Here's what typically goes inside it:

  • FROM: This directive specifies the base image to use for the Docker container. For example, we use buildpack-deps:bookworm as our starting point.
  • RUN: This command executes during the building phase of the container. With RUN, you can install packages, copy files from your computer to the container, or perform other setup tasks.
  • ENV: This sets environment variables within the container, similar to how the env command works in Linux. It can define things like the location of Python, your username, and more.
  • CMD: This specifies the command that runs when the container starts. It acts as the default entry point of the container. In our example, the container starts with CMD ["python3"]. The syntax CMD ["executable", "parameter1", "parameter2", ...] mirrors how you would run a command in a terminal, such as python3 myscript.py -i inputfile.txt.

These elements together define how the Docker container is built and what it does when it runs.

Pull Python image from DockerHub

The actual term for "install" is pull. As mentioned above, I randomly select python continer "bookworm 3.9". Let's pull the python docker image with a name specification to the computer. During the execution

docker pull python:3.9-bookworm  # semicolon : allows to specify version tag

you will see pulling The resulting image can be found in the docker images

docker images  # list all images stored in the computer
REPOSITORY   TAG                  IMAGE ID       CREATED        SIZE
python       3.10.14-alpine3.20   f484900e23bd   8 days ago     55.7MB
python       3.9-bookworm         9dcab61c255c   3 months ago   997MB

You see 3.10.14-alpine3.20 because I tried to pull other version as well. To delete it run

#docker rmi IMAGE_ID # rm remove, i image
docker rmi f484900e23bd  # image_id is  f484900e23bd

Run the container

Docker creates a container from image and from the same image one can create many containers. To see available docker images on the computer, run

docker images

which outputs

REPOSITORY   TAG                  IMAGE ID       CREATED        SIZE
python       3.9-bookworm         9dcab61c255c   3 months ago   997MB

We want to run this python repositpry. It can be done in 2 similar ways -- with image ID or with repository name + tag

docker run python:3.9-bookworm
# OR equivalently
# docker run 9dcab61c255c

Executing docker run python:3.9-bookworm did nothing. But why? Docker opened python successfully, didn't see anything to execute and exited. But we want to have an access to the python interactive terminal.

docker run -it python:3.9-bookworm  # -it is interactive terminal

gives the desired result. Great! How to install now numpy ? Or how can I I navige in folders of a new container? We do that with terminal.

As we discussed above, the CMD ["python3"] in Dockerfile configurations defines a default entrypoint which is python3. But we can change it easily.

Running the following command will start a container and open a Bash shell:

docker run -it python:3.9-bookworm bash 
# OR
# docker run -it python:3.9-bookworm /bin/bash
# OR
# docker run -it python:3.9-bookworm /bin/sh  # for shell if usual bash is not installed

Once inside the container, you should see a prompt like this:

root@3978e2e905d1:/#

In this mode, all standard terminal commands work fine. For instance, running whoami will confirm that you are logged in as root. You can navigate the virtual file system of the container just as if you were logged into an external computer.

To exit from the container terminal execute exit in containers bash.

We just created a docker container from the docker image. As I mentioned above, one image can be used to build many containers and actual docker run -it container_name does that.

Which containers were built? To see the list of them run

docker ps -a # ps is process status, -a is all
# docker ps  # lists all the active containers running on background

On my computer it returns

CONTAINER ID   IMAGE                 COMMAND     CREATED             STATUS                           PORTS     NAMES
3978e2e905d1   python:3.9-bookworm   "bash"      About an hour ago   Exited (0) 8 minutes ago                   reverent_blackwell
1d1f582c6e53   python:3.9-bookworm   "python3"   About an hour ago   Exited (0) About an hour ago               adoring_austin
4e9a0105c449   python:3.9-bookworm   "python3"   About an hour ago   Exited (0) About an hour ago               fervent_elion
fe048a469382   9dcab61c255c          "bash"      About an hour ago   Exited (130) About an hour ago             cranky_chebyshev
0c20f55d774d   9dcab61c255c          "python3"   2 hours ago         Exited (0) 2 hours ago                     friendly_curie
54112e00fcf6   9dcab61c255c          "python3"   2 hours ago         Exited (0) 2 hours ago                     elastic_shamir
ced62372fa31   9dcab61c255c          "python3"   2 hours ago         Exited (0) 2 hours ago                     jovial_lichterman

The same image python:3.9-bookworm created several containers. One on the top was created with

docker run -it python:3.9-bookworm bash

and "COMMAND" indicates the entrypoint command that runs container. Others were executed without explicitly mentioning bash as the entrypoint

docker run -it python:3.9-bookworm  # not always the best practice

"COMMAND" shows python3 which is the default entrypoint, as I mentioned when discussing option CMD ["python3"] in the Dockerfile.

Tip:
Using the -it option allows you to reuse the container later. Without it, future attempts to start the container will result in it immediately closing. This is expected behavior because the computer starts Python in the container, receives no further commands, and exits successfully. Remember to include -it with docker run for interactive terminal containers.

All of the containers above, if not used, can be deleted after with

# docker rm container_id
docker rm 3978e2e905d1

or

# docker rm container_name
docker rm reverent_blackwell  # name are generated randomly. You can chose a name as well

A container can be removed automatically after exiting, if you provide docker run with a corresponding command --rm

docker run --rm -it python:3.9-bookworm bash  # --rm means remove after exiting

In contrast, containers that are not removed can be reused. To start the container

# docker start container_id
docker start 3978e2e905d1  # container_id can be obtained from docker ps -a

access to bash

# docker exec -it container_id Executable
# execute a non default command
docker exec -it 3978e2e905d1 bash  # or /bin/bash to get access to bash

or run a python script

# docker exec -it container_id Executable parameter1 parameter2
docker exec -it 3978e2e905d1 python /path_to/folder/my_script.py

Later, after use of the container, exit and stop it

#docker stop container_id
docker stop 3978e2e905d1

Now let's install numpy and run some calculations

Next page Docker page 3. Installing numpy

Previous page Docker page 1. Using Python in the container