Docker container image capable of running an iPython notebook server with ability to access an Oracle database.
You installed Git and Docker.
If you are on OS X or Linux, you may already have git installed (open a terminal and type the
command git --version; if it returns a version number, you've already got git).
First, clone a copy of this repository:
git clone https://github.com/bmamlin/docker-ipython-oracle
This will make a docker-ipython-oracle folder with a copy of the contents of this repository.
For this build to work, you will need to supply your own RPMs for Oracle's Instant Client Basic Lite and SDK packages for Linux x86-64. These are only available once to you once you have registered an Oracle account and agreed to their license agreement.
-
If you haven't already, register for an Oracle account.
-
Once logged in to Oracle's website, go to Instant Client Downloads for Linux x86-64. You will need to accept their license agreement before you can download anything.
-
Pay close attention to version number and file names. For this repository, I am using version 11.2. You want to download two file: Basic Lite and SDK RPMs (file name should end with .rpm).
-
Copy the two Oracle RPM files into the rpms folder.
NOTE: This docker image is built assuming Oracle version 11.2. If you want to use IPython to connect to a different version of Oracle, then you'll want to use the corresponding version and will need to edit the three other places in the Docker file where the versino ("11.2") is mentioned to match your version.
Within the folder containing Dockerfile, type this command:
docker build -t ipython .
This is the longest step (can take 5-10 minutes or more depending on your computer and your
internet connection. You'll see a lot of stuff scrolling by and a few obscure, python-related
error messages that can be ignored. Ideally, it will end with a message saying that the image
was successfully built. Lines like Successfully installed cx-oracle are reassuring to see as
well. If it fails to build, look at the output to try to sort out the problem. Did it fail to
find either of your RPMs? Use docker ps -a to list out any failed containers and remove them
with docker rm ID_OR_NAME. Use docker images to list any images. You can leave the ubuntu
image (so it doesn't need to get downloaded again), but you can remove any unnamed images with
docker rmi ID. Note that you only need to type the first few characters in an ID for docker
to unambiguously recognize it.
Type this command:
docker run -d -p 80:8080 -p 8022:22 --name ipython ipython
The -d tells docker to run as a background daemon, so you return immediately to the command
line. The -p 80:8080 tells docker to map port 8080 (exposed by the image, see Dockerfile)
as port 80 on the host. '-p8022:22' tells docker to map port 22 (for ssh access) as port 8022
on the host. --name ipython gives the container a convenient name. And the ipython at the
end refers the image you built in Step Three.
You should now be able to use your IPython Notebook. If you are in Windoze or Mac using
boot2docker, then use the command boot2docker ip
to find the IP address of your virutal host machine (usually something like 192.168.59.103, but
ymmv). Assuming you find IPython waiting for you at that address, you can use a recipe like
this to reach your Oracle database:
%load_ext sql
%sql oracle://username:password@hostname:1521/sid
# Assuming that got you connected, you can start issuing SQL commands like
%sql SELECT * FROM PRODUCT_COMPONENT_VERSION
If you need to perform updates, you can ssh into the IPython image using username root and password ipython:
ssh -p 8022 root@192.168.59.103
# password is: ipython
Docker containers are maintained in memory and ephemeral. Anything you create in your notebook that you want to keep should be exported and saved. If your container is stopped for any reason (turning off docker for a while or rebooting your computer) you may find that you have an empty notebook again. You've been warned.