Dockerfile.spark

FROM centos:7

# Build ARGs
# Forcing back to python 3.8 to be in sync with local dev env.
# Can't run driver and worker on different Python versions when driver is local dev machine and not spark-submit container
ARG PYTHON_VERSION=3.10.12
ARG SPARK_VERSION=3.5.0
ARG HADOOP_VERSION=3.3.4
ARG PROJECT_LOG_DIR=/logs

RUN yum -y update && yum clean all
# sqlite-devel added as prerequisite for coverage python lib, used by pytest-cov plugin
RUN yum -y install wget gcc openssl-devel bzip2-devel libffi libffi-devel zlib-devel sqlite-devel
RUN yum -y groupinstall "Development Tools"

# Building Python 3.x
WORKDIR /usr/src
RUN wget --quiet https://www.python.org/ftp/python/${PYTHON_VERSION}/Python-${PYTHON_VERSION}.tgz \
    && tar xzf Python-${PYTHON_VERSION}.tgz
WORKDIR /usr/src/Python-${PYTHON_VERSION}
RUN ./configure --enable-optimizations \
    && make altinstall \
    && ln -sf /usr/local/bin/python`echo ${PYTHON_VERSION} | awk -F. '{short_version=$1 FS $2; print short_version}'` /usr/bin/python3 \
    && echo "Installed $(python3 --version)"
# Ensure Python STDOUT gets sent to container logs
ENV PYTHONUNBUFFERED=1

# Install Java 1.8.x
RUN yum -y install java-1.8.0-openjdk
ENV JAVA_HOME=/usr/lib/jvm/jre

# Install Hadoop and Spark
WORKDIR /usr/local
RUN wget --quiet https://archive.apache.org/dist/hadoop/common/hadoop-${HADOOP_VERSION}/hadoop-${HADOOP_VERSION}.tar.gz \
    && tar xzf hadoop-${HADOOP_VERSION}.tar.gz \
    && ln -sfn /usr/local/hadoop-${HADOOP_VERSION} /usr/local/hadoop \
    && wget --quiet https://archive.apache.org/dist/spark/spark-${SPARK_VERSION}/spark-${SPARK_VERSION}-bin-without-hadoop.tgz \
    && tar xzf spark-${SPARK_VERSION}-bin-without-hadoop.tgz \
    && ln -sfn /usr/local/spark-${SPARK_VERSION}-bin-without-hadoop /usr/local/spark \
    && echo "Installed $(/usr/local/hadoop/bin/hadoop version)"
ENV HADOOP_HOME=/usr/local/hadoop
ENV SPARK_HOME=/usr/local/spark
# Cannot set ENV var = command-result, [i.e. doing: ENV SPARK_DIST_CLASSPATH=$(${HADOOP_HOME}/bin/hadoop classpath)], so interpolating the hadoop classpath the long way
ENV SPARK_DIST_CLASSPATH="$HADOOP_HOME/etc/hadoop/*:$HADOOP_HOME/share/hadoop/common/lib/*:$HADOOP_HOME/share/hadoop/common/*:$HADOOP_HOME/share/hadoop/hdfs/*:$HADOOP_HOME/share/hadoop/hdfs/lib/*:$HADOOP_HOME/share/hadoop/hdfs/*:$HADOOP_HOME/share/hadoop/yarn/lib/*:$HADOOP_HOME/share/hadoop/yarn/*:$HADOOP_HOME/share/hadoop/mapreduce/lib/*:$HADOOP_HOME/share/hadoop/mapreduce/*:$HADOOP_HOME/share/hadoop/tools/lib/*"
ENV PATH=${SPARK_HOME}/bin:${HADOOP_HOME}/bin:${JAVA_HOME}/bin:${PATH};
RUN echo "Installed Spark" && echo "$(${SPARK_HOME}/bin/pyspark --version)"

# Config for starting up the Spark History Server
# This allows a web entrypoint to see past job runs, and bring up their Spark UI
# It allows job-run event logs to be monitored (e.g. at /tmp/spark-events)
# NOTE: The specified eventlog dir and its path does not exist. It is assumed that a host dir will be bind-mounted at /project
#       so that all containers that do the same can share the events.
RUN cp $SPARK_HOME/conf/spark-defaults.conf.template $SPARK_HOME/conf/spark-defaults.conf \
    && sed -i "s|^# spark.eventLog.dir.*$|spark.eventLog.dir                 file:///project/$PROJECT_LOG_DIR/spark-events|g" $SPARK_HOME/conf/spark-defaults.conf \
    && sed -i '/spark.eventLog.enabled/s/^# //g' $SPARK_HOME/conf/spark-defaults.conf \
    && echo "spark.history.fs.logDirectory      file:///project/$PROJECT_LOG_DIR/spark-events" >> $SPARK_HOME/conf/spark-defaults.conf

# Bake project dependencies into any spark-base -based container that will run Project python code (and require these dependencies)
# NOTE: We must exclude any dependencies on the host machine installed in a virtual environment (assumed to be stored under ./.venv)
WORKDIR /project
COPY requirements/ /project/requirements
RUN rm -rf /project/.venv
RUN python3 -m pip install -r requirements/requirements.txt

# Allow for entrypoints and commands to infer that they are running relative to this directory
WORKDIR ${SPARK_HOME}