Add Dockerfile.

Author: shimamoto

Dockerfile

@@ -0,0 +1,43 @@
+FROM ubuntu
+MAINTAINER shimamoto
+
+ENV PIO_VERSION 0.11.0-v1-SNAPSHOT
+ENV SPARK_VERSION 1.6.3
+ENV ELASTICSEARCH_VERSION 1.7.6
+ENV HBASE_VERSION 1.0.0
+
+ENV PIO_HOME /PredictionIO-${PIO_VERSION}
+ENV PATH=${PIO_HOME}/bin:$PATH
+ENV JAVA_HOME /usr/lib/jvm/java-8-openjdk-amd64
+
+RUN apt-get update \
+    && apt-get install -y --auto-remove --no-install-recommends curl git openjdk-8-jdk libgfortran3 python-pip \
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN cd / \
+    && git clone https://github.com/jpioug/incubator-predictionio.git \
+    && cd incubator-predictionio \
+    && ./make-distribution.sh
+
+RUN tar zxvf /incubator-predictionio/PredictionIO-${PIO_VERSION}.tar.gz -C /
+RUN rm -r /incubator-predictionio
+RUN mkdir /${PIO_HOME}/vendors
+COPY files/pio-env.sh ${PIO_HOME}/conf/pio-env.sh
+
+RUN curl -O http://d3kbcqa49mib13.cloudfront.net/spark-${SPARK_VERSION}-bin-hadoop2.6.tgz \
+    && tar -xvzf spark-${SPARK_VERSION}-bin-hadoop2.6.tgz -C ${PIO_HOME}/vendors \
+    && rm spark-${SPARK_VERSION}-bin-hadoop2.6.tgz
+
+RUN curl -O https://download.elasticsearch.org/elasticsearch/elasticsearch/elasticsearch-${ELASTICSEARCH_VERSION}.tar.gz \
+    && tar -xvzf elasticsearch-${ELASTICSEARCH_VERSION}.tar.gz -C ${PIO_HOME}/vendors \
+    && rm elasticsearch-${ELASTICSEARCH_VERSION}.tar.gz \
+    && echo 'cluster.name: predictionio' >> ${PIO_HOME}/vendors/elasticsearch-${ELASTICSEARCH_VERSION}/config/elasticsearch.yml \
+    && echo 'network.host: 127.0.0.1' >> ${PIO_HOME}/vendors/elasticsearch-${ELASTICSEARCH_VERSION}/config/elasticsearch.yml
+
+RUN curl -O http://archive.apache.org/dist/hbase/hbase-${HBASE_VERSION}/hbase-${HBASE_VERSION}-bin.tar.gz \
+    && tar -xvzf hbase-${HBASE_VERSION}-bin.tar.gz -C ${PIO_HOME}/vendors \
+    && rm hbase-${HBASE_VERSION}-bin.tar.gz
+COPY files/hbase-site.xml ${PIO_HOME}/vendors/hbase-${HBASE_VERSION}/conf/hbase-site.xml
+RUN sed -i "s|VAR_PIO_HOME|${PIO_HOME}|" ${PIO_HOME}/vendors/hbase-${HBASE_VERSION}/conf/hbase-site.xml \
+    && sed -i "s|VAR_HBASE_VERSION|${HBASE_VERSION}|" ${PIO_HOME}/vendors/hbase-${HBASE_VERSION}/conf/hbase-site.xml

README.md

@@ -1,2 +1,21 @@
 # docker-predictionio
 Docker container for PredictionIO-based machine learning services
+
+This container uses Apache Spark, HBase and Elasticsearch.
+The PredictionIO version is [jpioug](https://github.com/jpioug) version.
+
+### Use it interactively for development:
+1. First, build docker image from local Dockerfile: cd to the path containing the Dockerfile, then:
+```Bash
+$ docker build -t predictionio .
+```
+then:
+```Bash
+$ docker run -p 8000:8000 --name predictionio_instance -it predictionio /bin/bash
+```
+
+2. Then in docker container, start all services and check they are started
+```Bash
+$ pio-start-all
+$ jps -l
+```

files/hbase-site.xml

@@ -0,0 +1,10 @@
+<configuration>
+  <property>
+    <name>hbase.rootdir</name>
+    <value>file://VAR_PIO_HOME/vendors/hbase-VAR_HBASE_VERSION/data</value>
+  </property>
+  <property>
+    <name>hbase.zookeeper.property.dataDir</name>
+    <value>VAR_PIO_HOME/vendors/zookeeper</value>
+  </property>
+</configuration>

files/pio-env.sh

@@ -0,0 +1,26 @@
+#!/usr/bin/env bash
+
+PIO_FS_BASEDIR=${HOME}/.pio_store
+PIO_FS_ENGINESDIR=${PIO_FS_BASEDIR}/engines
+PIO_FS_TMPDIR=${PIO_FS_BASEDIR}/tmp
+
+SPARK_HOME=${PIO_HOME}/vendors/spark-${SPARK_VERSION}-bin-hadoop2.6
+
+HBASE_CONF_DIR=${PIO_HOME}/vendors/hbase-${HBASE_VERSION}/conf
+
+PIO_STORAGE_REPOSITORIES_METADATA_NAME=pio_meta
+PIO_STORAGE_REPOSITORIES_METADATA_SOURCE=ELASTICSEARCH
+PIO_STORAGE_REPOSITORIES_EVENTDATA_NAME=pio_event
+PIO_STORAGE_REPOSITORIES_EVENTDATA_SOURCE=HBASE
+PIO_STORAGE_REPOSITORIES_MODELDATA_NAME=pio_model
+PIO_STORAGE_REPOSITORIES_MODELDATA_SOURCE=LOCALFS
+
+PIO_STORAGE_SOURCES_ELASTICSEARCH_TYPE=elasticsearch
+PIO_STORAGE_SOURCES_ELASTICSEARCH_CLUSTERNAME=predictionio
+PIO_STORAGE_SOURCES_ELASTICSEARCH_HOSTS=localhost
+PIO_STORAGE_SOURCES_ELASTICSEARCH_PORTS=9300
+PIO_STORAGE_SOURCES_ELASTICSEARCH_HOME=${PIO_HOME}/vendors/elasticsearch-${ELASTICSEARCH_VERSION}
+PIO_STORAGE_SOURCES_LOCALFS_TYPE=localfs
+PIO_STORAGE_SOURCES_LOCALFS_PATH=${PIO_FS_BASEDIR}/models
+PIO_STORAGE_SOURCES_HBASE_TYPE=hbase
+PIO_STORAGE_SOURCES_HBASE_HOME=${PIO_HOME}/vendors/hbase-${HBASE_VERSION}