HDFS高可用環境搭建

                             HA(high availability）

    HA 使用的是分佈式日誌管理方式

    1.  問題
    Namenode出現問題，整個集羣將不能使用。
    配置兩個namenode:Active namenode，standby namenode

    2.  實現方式
    1.  兩個namenode內存中存儲的元數據同步，namenode啓動時，會讀鏡像文件。
    2.  編輯日誌的安全
    分佈式的存儲日誌文件，存儲於2n+1奇數個節點。（n個節點寫入成功，日誌寫入成功。）
    Zookeeper監控
    監控兩個namenode，一個namenode出現問題，實現故障轉移。
    Zookeeper對時間同步要求較高（ntp時間同步）
    3.  客戶端如何知道訪問哪一個namenode

    使用proxy代理
    隔離機制
    使用sshfence
    兩個namenode之間無密碼登陸

    安裝配置
    1.  基礎環境配置

node1	node2	node3	node1	node2
192.168.103.26	192.168.103.27	192.168.103.28	192.168.103.29	192.168.103.30
namenode	namenode	datanode	datanode	datanode
DFSZKFailoverController	DFSZKFailoverController	journalnode	journalnode	journalnode
		QuorumPeerMain	QuorumPeerMain	QuorumPeerMain

    配置主機名與IP之間的映射
    vim /etc/hosts
    192.168.103.26 node1
    192.168.103.27 node2
    192.168.103.28 node3
    192.168.103.29 node4
    192.168.103.30 node5

    配置各個節點之間的免密登陸
    Node1
    ssh-kengen –t rsa –P ‘’ 在~/.ssh/目錄下生成id_rsa, id_rsa.put密鑰
    ssh-copy-id –I ~/.ssh/id._rsa.pub (node1,node2,node3,node4.node5)
    Node2操作同node1

    配置時間同步,node1作爲ntp服務器
    1．  yum install ntp –y （所有節點）

    2．  node1
            vim /etc/ntp.conf
    server 210.72.145.44    # 中國國家受時中心
    server    127.127.1.0     #爲局域網用戶提供服務
    fudge     127.127.1.0     stratum 10
    systemctl start ntpd
    3.node2,node3,node4,node5
    ntpdate node1

    2.  安裝hadoop

    1.  tar –zxvf jdk-8u171-linux-x64.tar.gz –C /
    mv jdk1.8.0_171/ jdk
    tar –zxvf hadoop-2.7.7.tar.gz –C /
    mv hadoop-2.7.7/ Hadoop
    tar –zxvf zookeeper-3.4.10.tar.gz –C /
    mv zookeeper-3.4.10 zookeeper

    3.  vim /etc/profile
    export JAVA_HOME=/jdk
    export HADOOP_HOME=/Hadoop
    export ZOOKEEPER_HOME=/zookeeper
    export PATH=$PATH:$JAVA_HOME/bin:$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$ZOOKEEPER_HOME/bin
    source /etc/profile
    scp /etc/profile node2:/etc/
             node3,node4,node5
    scp –r /jdk node2:/etc
         node3,node4,node5  

    4.  配置zookeeper
    進入zookeeper目錄，創建zkdata目錄
    創建myid文件，node3,node4,node5文件中的值爲1，2，3
    scp –r /zookeeper node2:/etc
         node3,node4,node5
分別修改zookeeper節點的myid文件

    5.  安裝hadoop(重點!!!)
    1.  hadoop-env.sh
    export JAVA_HOME = /jdk

    2.  core-site.xml
    <configuration>
    <property>
    <name>fs.defaultFS</name>
    <value>hdfs://ns1</value>
    </property>
    <property>
    <name>ha.zookeeper.quorum</name>
    <value>node3:2181,node4:2181,node5:2181</value>
</property>
    <property>
    <name>hadoop.tmp.dir</name>
    <value>/hadoop/tmp</value>
</property>
    </configuration>

    3.  hdfs.site.xml
    <configuration>
    <property>
            <name>dfs.nameservices</name>
            <value>ns1</value>
    </property>
    <property>
            <name>dfs.ha.namenodes.ns1</name>
            <value>nn1,nn2</value>
    </property>
    <property>
            <name>dfs.namenode.rpc-address.ns1.nn1</name>
            <value>node1:8020</value>
    </property>
    <property>
            <name>dfs.namenode.rpc-address.ns1.nn2</name>
            <value>node2:8020</value>
            </property>
    <property>
            <name>dfs.ha.namenode.http-address.ns1.nn1</name>
            <value>node1:50070</value>
    </property>
    <property>
            <name>dfs.ha.namenode.http-address.ns1.nn2</name>
            <value>node2:50070</value>
    </property>
    <property>
            <name>dfs.namenode.shared.edits.dir</name>
            <value>qjournal://node3:8485;node4:8485;node5:8485/ns1</value>
    </property>
    <property>
            <name>dfs.journalnode.edits.dir</name>
            <value>/journalnode</value>
    </property>
    <property>
            <name>dfs.client.failover.proxy.provider.ns1</name>
    <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
    </property>
    <property>
            <name>dfs.ha.fencing.methods</name>
            <value>sshfence</value>
    </property>
    <property>
            <name>dfs.ha.fencing.ssh.private-key-files</name>
            <value>/home/root/.ssh/id_rsa</value>
        </property>
<property>
            <name>dfs.ha.automatic-failover.enabled</name>
            <value>true</value>
</property>

    vim slaves
    node3
    node4
    node5

    4.  啓動
    node3,node4,node5
    hadoop-daemon.sh start journalnode
    zkServer.sh start
  node1
    hdfs namenode -format
        scp –r /Hadoop/tmp node2:/Hadoop/
  hdfs zkfc –formatZK
  start-dfs.sh

    5.  驗證HDFS HA
    通過瀏覽器查看node1與node2 namenode狀態
    ![](https://s1.51cto.com/images/blog/201901/14/bc5b4076ff24c413c5e95455950bbfcc.png?x-oss-process=image/watermark,size_16,text_QDUxQ1RP5Y2a5a6i,color_FFFFFF,t_100,g_se,x_10,y_10,shadow_90,type_ZmFuZ3poZW5naGVpdGk=)
    ![](https://s1.51cto.com/images/blog/201901/14/28b7afefc3ed4c7556b6611bdb1fb328.png?x-oss-process=image/watermark,size_16,text_QDUxQ1RP5Y2a5a6i,color_FFFFFF,t_100,g_se,x_10,y_10,shadow_90,type_ZmFuZ3poZW5naGVpdGk=)

    hadoop-daemon.sh stop namenode

    ![](https://s1.51cto.com/images/blog/201901/14/b23438b0ec3b16c3ca045e7b5bd73282.png?x-oss-process=image/watermark,size_16,text_QDUxQ1RP5Y2a5a6i,color_FFFFFF,t_100,g_se,x_10,y_10,shadow_90,type_ZmFuZ3poZW5naGVpdGk=)