install_hadoop.yml

# ansible-playbook -i hosts install_hadoop.yml
# 设置hostname
# ssh-keygen 并互相免密登录
# ./bin/hdfs namenode -format
# ./sbin/start-all.sh
# TODO 删除block marker
- hosts: hadoop
  vars:
    target: /opt/soft
    unzip_dir: hadoop-3.2.0
    tarball: hadoop-3.2.0.tar.gz

  tasks:
    - name: gather JAVA_HOME
      shell: source /etc/profile && echo $JAVA_HOME
      register: JAVA_HOME

    - name: Install Package
      yum:
        name:
          - vim
          - ntp
          - telnet

#    - name: every minute
#      cron:
#        minute: "*/1"
#        job: "/usr/sbin/ntpdate -u ntp.aliyun.com"

    - name: stop and disable firewalld.
      service:
        name: firewalld
        state: stopped
        enabled: False

    - name: set hosts
      shell: /bin/echo {{ item }} >> /etc/hosts
      with_items:
        - 192.168.216.149 hadoop01
        - 192.168.216.148 hadoop02
        - 192.168.216.150 hadoop03
    - name: copy hadoop remote hosts
      copy: src=/root/{{ tarball }} dest={{ target }}

    - name: Decompressing files
      shell: chdir={{ target }} tar zxvf {{ tarball }}

    - name: hadoop_profile config
      shell: /bin/echo {{ item }} >> /etc/profile
      with_items:
        - \# Hadoop Env
        - export HADOOP_HOME={{ target }}/{{ unzip_dir }}
        - export PATH=\$PATH:\$HADOOP_HOME/bin
        - export PATH=\$PATH:\$HADOOP_HOME/sbin
    - name: take effect
      shell: source /etc/profile
    - name: hadoop-env config
      shell: /bin/echo {{ item }} >> {{ target }}/{{ unzip_dir }}/etc/hadoop/hadoop-env.sh
      with_items:
        - export JAVA_HOME={{ JAVA_HOME.stdout }}
        - export HADOOP_LOG_DIR=/data/hadoop_repo/logs/hadoop
    - name: modify core-site.xml
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/etc/hadoop/core-site.xml"
        block: |
          <property>
              <name>fs.defaultFS</name>
              <value>hdfs://hadoop01:9000</value>
          </property>
          <property>
              <name>hadoop.tmp.dir</name>
              <value>/data/hadoop_repo</value>
          </property>
          <!--回收站-->
          <property>
            <name>fs.trash.interval</name>
            <value>1440</value>
          </property>

        insertafter: "<configuration>"
    - name: modify hdfs-site.xml
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/etc/hadoop/hdfs-site.xml"
        block: |
          <property>
              <name>dfs.replication</name>
              <value>2</value>
          </property>
          <property>
              <name>dfs.namenode.secondary.http-address</name>
              <value>hadoop01:50090</value>
          </property>
          <property>
              <name>dfs.namenode.datanode.registration.ip-hostname-check</name>
              <value>false</value>
          </property>
          <!--关闭权限-->
          <property>
            <name>dfs.permissions.enabled</name>
            <value>false</value>
          </property>
        insertafter: "<configuration>"
    - name: modify mapred-site.xml
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/etc/hadoop/mapred-site.xml"
        block: |
          <property>
              <name>mapreduce.framework.name</name>
              <value>yarn</value>
          </property>
        insertafter: "<configuration>"
    - name: modify yarn-site.xml
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/etc/hadoop/yarn-site.xml"
        block: |
          <property>
              <name>yarn.nodemanager.aux-services</name>
              <value>mapreduce_shuffle</value>
          </property>
          <property>
              <name>yarn.nodemanager.env-whitelist</name>
              <value>
                  JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,CLASSPATH_PREPEND_DISTCACHE,HADOOP_YARN_HOME,HADOOP_MAPRED_HOME
              </value>
          </property>
          <property>
              <name>yarn.resourcemanager.hostname</name>
              <value>hadoop01</value>
          </property>
        insertafter: "<configuration>"

    - name: modifiy workers
      replace:
        path: "{{ target }}/{{ unzip_dir }}/etc/hadoop/workers"
        regexp: "localhost"
        replace: |
          hadoop02
          hadoop03

    - name: modify start-dfs.sh
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/sbin/start-dfs.sh"
        block: |
          HDFS_DATANODE_USER=root
          HDFS_DATANODE_SECURE_USER=hdfs
          HDFS_NAMENODE_USER=root
          HDFS_SECONDARYNAMENODE_USER=root

        insertafter: "#!/usr/bin/env bash"
    - name: modify stop-dfs.sh
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/sbin/stop-dfs.sh"
        block: |
          HDFS_DATANODE_USER=root
          HDFS_DATANODE_SECURE_USER=hdfs
          HDFS_NAMENODE_USER=root
          HDFS_SECONDARYNAMENODE_USER=root

        insertafter: "#!/usr/bin/env bash"
    - name: modify start-yarn.sh
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/sbin/start-yarn.sh"
        block: |
          YARN_RESOURCEMANAGER_USER=root
          HADOOP_SECURE_DN_USER=yarn
          YARN_NODEMANAGER_USER=root
        insertafter: "#!/usr/bin/env bash"
    - name: modify stop-yarn.sh
      blockinfile:
        path: "{{ target }}/{{ unzip_dir }}/sbin/stop-yarn.sh"
        block: |
          YARN_RESOURCEMANAGER_USER=root
          HADOOP_SECURE_DN_USER=yarn
          YARN_NODEMANAGER_USER=root
        insertafter: "#!/usr/bin/env bash"
    - name: clean tarball
      file:
        path: "{{ target }}/{{ tarball }}"
        state: absent