tree: 97494984c6bdddc45d522dca472ffc36544aa428 [path history] [tgz]
  1. predictionio/
  2. spark/
  3. postgresql.yaml
  4. predictionio_postgresql.yaml
  5. README.md
docker/charts/README.md

Helm Charts for Apache PredictionIO

Overview

Helm Charts are packages of pre-configured Kubernetes resources. Using charts, you can install and manage PredictionIO in the Kubernetes.

Usage

Install PredictionIO with PostgreSQL

To install PostgreSQL and PredictionIO, run helm install command:

helm install --name my-postgresql stable/postgresql -f postgresql.yaml
helm install --name my-pio ./predictionio -f predictionio_postgresql.yaml

postgresql.yaml and predictionio_postgresql.yaml are configuration files for charts. To access Jupyter for PredictionIO, run kubectl port-forward and then open http://localhost:8888/.

export POD_NAME=$(kubectl get pods --namespace default -l "app.kubernetes.io/name=predictionio,app.kubernetes.io/instance=my-pio" -o jsonpath="{.items[0].metadata.name}")
kubectl port-forward $POD_NAME 8888:8888

Install Spark Cluster

To install Spark cluster, run the following command:

helm install --name my-spark ./spark

To train a model, run pio train as below:

pio train -- --master spark://my-spark-master:7077