Deiteo (데이터) means data in korean. This project builds up an environment consisting of, spark, kubernetes.

This project sets up a k8s cluster and deploys a spark docker container within the cluster with a master and a configurable amount of workers.

Official spark doc

The project has an example application that can be run locally and from within the cluster.

The main idea however with this repository, is to create a k8s spark cluster which you can deploy spark applications against, either pyspark or scala applications.

• poetry
  • curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python -
  • source $HOME/.poetry/env
• docker
• python3.9.1
• kubectl
• minikube

Project uses sonarcloud to measure quality.

For a fresh initial install, depending on the pyproject.lock:

make build

If a new dependency is added, or removed, the pyproject.lock needs to be updated:

make update

Install kubectl:

brew install kubectl

Test to ensure the version you installed is up-to-date:

kubectl version --client

Install minikube:

brew install minikube

Note that you will need at least PyCharm version 2020.3.

Install Plugin to IDE Here!

Enter:

Preferences -> Python Interpreter -> Add -> Poetry Environment -> Existing Environment (`if you ran make build it will automatically find it`)

To see what specific configuration that has been made to the pytest test, please look at:

.coveragec and pytest.ini, for more info read:

• pytest
• pytest-cov with .coveragec

Example output:

Name                                                    Stmts   Miss  Cover   Missing
-------------------------------------------------------------------------------------
test_0                                                   43      3    93%   34-35, 39
. . .
test_N                                                   51     39    24%   21-30
-------------------------------------------------------------------------------------
TOTAL                                                    94     43    51%

• Stmts - Total lines of code in a specific file
• Miss - Total number of lines that are not covered
• Cover - Percentage of all line of code that are covered, or (Stmts - Miss) / 100
• Missing - Lines of codes that are not covered

To modify the .coverage file for unit then modify tests/coverage/unit/.coveragerc or tests/coverage/integration/.coveragerc.

They are separated because you might wanna exclude specific code via the omit based on if unit is integration are running.

make test type=unit

make test type=integration

If you want to run with debug mode with ipdb directly from the console:

Example:

make test type=unit test_argument=-s

class Something:
    def __init__(self):
        import ipdb
        ipdb.set_trace()
        self.foo = "foo"

• n to step forward
• s to step into
• type any variable name while stepping to inspect it

Please enter the link here: spark deployment on k8s to proceed with setting up spark on k8s needed for this project.