Build docker image with hops hadoop distribution

deploy/kubernetes/docker/Dockerfile

@@ -24,6 +24,7 @@ ARG RSS_VERSION
 ARG AUTHOR=unknown
 ARG GIT_BRANCH=unknown
 ARG GIT_COMMIT=unknown
+ARG HADOOP_FILE=hops-$HADOOP_VERSION.tgz
 
 LABEL author=${AUTHOR}
 LABEL git-branch=${GIT_BRANCH}
@@ -44,10 +45,10 @@ COPY rss-env.sh /data/rssadmin/rss/bin
 
 COPY start.sh /data/rssadmin/rss/bin
 
-COPY hadoop-${HADOOP_VERSION}.tar.gz /data/rssadmin
-RUN tar -zxvf /data/rssadmin/hadoop-${HADOOP_VERSION}.tar.gz -C /data/rssadmin
-RUN mv /data/rssadmin/hadoop-${HADOOP_VERSION} /data/rssadmin/hadoop
-RUN rm -rf /data/rssadmin/hadoop-${HADOOP_VERSION}.tar.gz
+COPY $HADOOP_FILE /data/rssadmin
+RUN tar -zxvf /data/rssadmin/$HADOOP_FILE -C /data/rssadmin
+RUN mv /data/rssadmin/hops-${HADOOP_VERSION} /data/rssadmin/hadoop
+RUN rm -rf /data/rssadmin/$HADOOP_FILE
 COPY hadoopconfig/ /data/rssadmin/hadoop/etc/hadoop
 
 ENV RSS_VERSION ${RSS_VERSION}

deploy/kubernetes/docker/build.sh

@@ -36,7 +36,6 @@ function exit_with_usage() {
 
 REGISTRY="docker.io/library"
 HADOOP_VERSION=2.8.5
-HADOOP_SHORT_VERSION=$(echo $HADOOP_VERSION | awk -F "." '{print $1"."$2}')
 AUTHOR=$(whoami)
 # If you are based in China, you could pass --apache-mirror <a_mirror_url> when building this.
 APACHE_MIRROR="https://dlcdn.apache.org"
@@ -92,6 +91,8 @@ while (( "$#" )); do
   shift
 done
 
+HADOOP_SHORT_VERSION=$(echo $HADOOP_VERSION | awk -F "." '{print $1"."$2}')
+
 if [ -z "$BASE_IMAGE" ]; then
   echo "start building base image: uniffle-base"
   docker build -t "uniffle-base:latest" \
@@ -101,14 +102,21 @@ else
   echo "using base image(${BASE_IMAGE}) to build rss server"
 fi
 
+if test -f "/root/.wgetrc"; then
+    wget https://nexus.hops.works/repository/hopshadoop/hops-$HADOOP_VERSION_EE.tgz
+else
+    wget https://repo.hops.works/master/hops-$HADOOP_VERSION.tgz
+fi
 
-HADOOP_FILE=hadoop-${HADOOP_VERSION}.tar.gz
-ARCHIVE_HADOOP_URL=https://archive.apache.org/dist/hadoop/core/hadoop-${HADOOP_VERSION}/${HADOOP_FILE}
-HADOOP_URL=${APACHE_MIRROR}/hadoop/core/hadoop-${HADOOP_VERSION}/${HADOOP_FILE}
-echo "HADOOP_URL is either ${HADOOP_URL} or ${ARCHIVE_HADOOP_URL}"
-if [ ! -e "$HADOOP_FILE" ]; \
-  then wget "${HADOOP_URL}" || wget "$ARCHIVE_HADOOP_URL"; \
-  else echo "${HADOOP_FILE} has been downloaded"; \
+HADOOP_FILE=hops-$HADOOP_VERSION.tgz
+if [ ! -e "$HADOOP_FILE" ]; then
+  if test -f "/root/.wgetrc"; then
+      wget https://nexus.hops.works/repository/hopshadoop/$HADOOP_FILE
+  else
+      wget https://repo.hops.works/master/$HADOOP_FILE
+  fi
+else
+  echo "${HADOOP_FILE} has been downloaded";
 fi
 
 RSS_DIR=../../..
@@ -137,6 +145,7 @@ docker build --network=host -t "$IMAGE" \
              --build-arg GIT_COMMIT="$GIT_COMMIT" \
              --build-arg GIT_BRANCH="$GIT_BRANCH" \
              --build-arg BASE_IMAGE="$BASE_IMAGE" \
+             --build-arg HADOOP_FILE="$HADOOP_FILE" \
              -f Dockerfile --no-cache .
 
 if [ x"${PUSH_IMAGE}" == x"true" ]; then