Blog

If you want to use Oracle Big Data Cloud and Python using the built-in Zeppelin notebooks, follow these instructions.  This is based on a single node instance.

Versions:

Anaconda 3.6 : 3.6.3

Oracle BDC: 17.3.3-20

Steps I took.

• Get Anaconda
  • login to BDC using ssh
  • mkdir anaconda
  • cd anaconda
  • wget https://repo.continuum.io/archive/Anaconda3-5.0.1-Linux-x86_64.sh
  • Or latest version
• Install Anaconda (I installed 3.6, 2.7 also works)
  • sudo bash Anaconda3-5.0.1-Linux-x86_64.sh
  • Accept Licence
  • install to /opt/anaconda
• Check if everything is OK
  • logout of ssh
  • connect to ssh again
  • /opt/anaconda/bin/conda list
  • If you get a list of packages, you are good to go.
• Modify Zeppelin interpreter in Oracle BDC to use Anaconda
  • Login to BDC console
  • Click on Settings tab.
  • Change/add the following properties
    • Change zeppelin.pyspark.python – Set to /opt/anaconda/bin/python
    • Add PYSPARK_DRIVER_PYTHON – Set to /opt/anaconda/bin/python
    • Add PYSPARK_PYTHON – Set to /opt/anaconda/bin/python
    • Click SAVE button
• Test the zeppelin interpreter
  • Go to the Notebook tab
  • Create a New Note
  • In first note type
    • %pyspark
      import sys
      print(sys.version)
  • If you see your version, you are all set.