update

hadoop/build.sh

@@ -7,6 +7,6 @@ docker stop hadoop
 docker rm hadoop
 
 docker run -dit --name hadoop --privileged=true \
-  -v /data/hadoop/dfs:/dfs/data \
+  --hostname hadoop01 \
   -p 8042 -p 8088 -p 19888 -p 50070 -p 50075 zeekling/hadoop
 

hadoop/hadoop/Dockerfile

@@ -33,7 +33,7 @@ RUN rm -rf /etc/yum.repos.d/*.repo
 
 COPY ./CentOS-8-reg.repo /etc/yum.repos.d/CentOS-Base.repo
 
-RUN yum clean all && yum makecache && yum install -y openssh-server openssh-clients tar which
+RUN yum clean all && yum makecache && yum install -y openssh-server openssh-clients tar which vim
 
 RUN set -eux && \
     yum install -y wget hostname && \
@@ -94,6 +94,6 @@ ENV YARN_NODEMANAGER_USER=yarn
 
 #EXPOSE 8020 8042 8088 9000 10020 19888 50010 50020 50070 50075 50090
 # Hadoop 3.0 changed ports :-(
-EXPOSE 8020 8042 8088 9000 9868 9870 10020 19888 50010 50020 50090
+EXPOSE 9870 9871 19888 8088 9864-9869 8020 9000 9820 8032
 
 CMD ["/entrypoint.sh"]

hadoop/hadoop/conf/core-site.xml

@@ -3,6 +3,6 @@
 <configuration>
     <property>
         <name>fs.defaultFS</name>
-        <value>hdfs://localhost:8020</value>
+        <value>hdfs://localhost:9820</value>
     </property>
 </configuration>

hadoop/hadoop/conf/hdfs-site.xml

@@ -5,6 +5,10 @@
     <name>dfs.replication</name>
     <value>1</value>
   </property>
+  <property>
+    <name>dfs.permissions</name>
+    <value>false</value>
+  </property>
   <property>
     <name>dfs.namenode.name.dir</name>
     <value>/dfs/name</value>

hadoop/hadoop/entrypoint.sh

@@ -99,6 +99,8 @@ EOF
     fi | tee -a /root/.ssh/known_hosts
 
     mkdir -pv /hadoop/logs
+    mkdir -p /dfs/data
+    chown -R hdfs:hadoop /dfs/
 
     sed -i "s/localhost/$hostname/" /hadoop/etc/hadoop/core-site.xml
     if [ -f "/run/nologin" ];then
@@ -107,6 +109,4 @@ EOF
     start-dfs.sh
     start-yarn.sh
     tail -f /dev/null /hadoop/logs/*
-    stop-yarn.sh
-    stop-dfs.sh
 fi