09-jupyter-and-spark2-connect.md

jupyter and spark2 connect

• there is only python3 kernel

jupyter kernelspec list

cat $HOME/$USER/delab-venv/share/jupyter/kernels/python3/kernel/json

• let's creat for pyspark2

mkdir $HOME/$USER/delab-venv/share/jupyter/kernels/pyspark2

vi $HOME/$USER/delab-venv/share/jupyter/kernels/pyspark2/kernel.json

{
      "argv": [
        "python",
        "-m",
        "ipykernel_launcher",
        "-f",
        "{connection_file}"
      ],
      "display_name": "Pyspark 2",
      "language": "python",
      "env": {
        "PYSPARK_PYTHON": "/usr/bin/python3",
        "SPARK_HOME": "/opt/spark2/",
        "SPARK_OPTS": "--master yarn --conf spark.ui.port=0",
        "PYTHONPATH": "/opt/spark2/python/lib/py4j-0.10.7-src.zip:/opt/spark2/python/"
      }
}

jupyter kernelspec install $HOME/$USER/delab-venv/share/jupyter/kernels/pyspark2 --user

• let's check

jupyter kernelspec list

• open jupyter lab ui on browser and you will see pyspark2 notebook.

from pyspark.sql import SparkSession

spark = SparkSession. \
    builder. \
    enableHiveSupport(). \
    appName('Demo'). \
    master('yarn'). \
    getOrCreate()

spark.sql('SHOW databases').show()

spark.sql('SELECT count(1) FROM retail_db.orders').show()