Hadoop2的HA安装(high availability):nfs+zookeeper

前面介绍过hadoop的简单安装和FA安装,在这里将介绍几种hadoop2中HA(高可用性)安装,HA技术使hadoop不再存在单点namenode的故障。

先来第一种:nfs+zookeeper

Hadoop 版本:2.2.0

OS 版本: Centos6.4

Jdk 版本: jdk1.6.0_32

环境配置

机器名

Ip地址

功能

Hadoop1

192.168.124.135

NameNode, DataNode,

ResourceManager, NodeManager

Zookeeper

Zkfc

Hadoop2

192.168.124.136

NameNode

DataNode, NodeManager

Zookeeper

Zkfc

Hadoop3

192.168.124.137

DataNode, NodeManager

Zookeeper

Zkfc

Nfs server

安装zookeeper

 使用 FileZilla上传zookeeper-3.4.5.tar.gz

  解压缩    tar xzvf zookeeper-3.4.5.tar.gz

配置zookeeper

Vi conf/zoo.cfg

# The number of milliseconds of each tick

tickTime=2000

# The number of ticks that the initial

# synchronization phase can take

initLimit=10

# The number of ticks that can pass between

# sending a request and getting an acknowledgement

syncLimit=5

# the directory where the snapshot is stored.

# do not use /tmp for storage, /tmp here is just

# example sakes.

dataDir=/home/hadoop/repo1/zookeeper

# the port at which the clients will connect

clientPort=2181

#

# Be sure to read the maintenance section of the

# administrator guide before turning on autopurge.

#

# http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance

#

# The number of snapshots to retain in dataDir

#autopurge.snapRetainCount=3

# Purge task interval in hours

# Set to "0" to disable auto purge feature

#autopurge.purgeInterval=1

 

server.1=hadoop1:2888:3888

server.2=hadoop2:2888:3888

server.3=hadoop3:2888:3888

 

在hadoop1, hadoop2, hadoop3, 修改 /home/hadoop/repo1/zookeeper/myid

按照下面的表哥填写myid

 

Hadoop1

1

Hadoop2

2

Hadoop3

3

nfs安装

在hadoop3上安装

yum install nfs-utils

vi /etc/exports

/home/hadoop/repo3/nfs 192.168.124.0/24(rw,sync,no_root_squash)

启动

service rpcbind restart

service nfs restart

在hadoop1和hadoop2运行mount命令

mount -t nfs hadoop3:/home/hadoop/repo3/nfs /home/hadoop/repo3/nfs

配置hadoop

vi etc/hadoop/hadoop-env.sh 修改jdk位置
export JAVA_HOME=/home/hadoop/jdk1.6.0_32

 

vi etc/hadoop/mapred-env.sh修改jdk位置

export JAVA_HOME=/home/hadoop/jdk1.6.0_32

 

vi etc/hadoop/yarn-env.sh修改jdk位置

export JAVA_HOME=/home/hadoop/jdk1.6.0_32

 

vi etc/hadoop/core-site.xml

<configuration>

    <property>

        <name>hadoop.tmp.dir</name>

        <value>/home/hadoop/repo3/tmp</value>

        <description>A base for other temporary directories.</description>

    </property>

    <property>

        <name>fs.defaultFS</name>

        <value>hdfs://mycluster</value>

    </property>

    <property>

        <name>dfs.journalnode.edits.dir</name>

        <value>/home/hadoop/repo3/journal</value>

    </property>

    <property>

        <name>ha.zookeeper.quorum</name>

        <value>hadoop1:2181,hadoop2:2181,hadoop3:2181</value>

    </property>

</configuration>

 

vi etc/hadoop/hdfs-site.xml

<configuration>

    <property>

        <name>dfs.replication</name>

        <value>2</value>

    </property>

    <property>

        <name>dfs.namenode.name.dir</name>

        <value>/home/hadoop/repo3/name</value>

    </property>

    <property>

        <name>dfs.datanode.data.dir</name>

        <value>/home/hadoop/repo3/data</value>

    </property>

    <property>

        <name>dfs.nameservices</name>

        <value>mycluster</value>

    </property>

    <property>

        <name>dfs.ha.namenodes.mycluster</name>

        <value>hadoop1,hadoop2</value>

    </property>

    <property>

        <name>dfs.namenode.rpc-address.mycluster.hadoop1</name>

        <value>hadoop1:9000</value>

    </property>

    <property>

        <name>dfs.namenode.http-address.mycluster.hadoop1</name>

        <value>hadoop1:50070</value>

    </property>

    <property>

        <name>dfs.namenode.rpc-address.mycluster.hadoop2</name>

        <value>hadoop2:9000</value>

    </property>

    <property>

        <name>dfs.namenode.http-address.mycluster.hadoop2</name>

        <value>hadoop2:50070</value>

    </property>

    <property>

        <name>dfs.namenode.shared.edits.dir</name>

        <value>file:///home/hadoop/repo3/nfs</value>

    </property>

    <property>

        <name>dfs.client.failover.proxy.provider.mycluster</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/hadoop/.ssh/id_rsa</value>

    </property>

    <property>

        <name>dfs.ha.automatic-failover.enabled</name>

        <value>true</value>

    </property>

</configuration>

 

vi etc/hadoop/yarn-site.xml

<configuration>

    <property>

        <description>the valid service name</description>

        <name>yarn.nodemanager.aux-services</name>

        <value>mapreduce_shuffle</value>

    </property>

    <property>

        <description>The hostname of the RM.</description>

        <name>yarn.resourcemanager.hostname</name>

        <value>hadoop1</value>

    </property>

</configuration>

 

vi etc/hadoop/mapred-site.xml

<configuration>

    <property>

        <name>mapreduce.framework.name</name>

        <value>yarn</value>

    </property>

</configuration>

 

vi etc/hadoop/slaves

hadoop1

hadoop2

hadoop3

 

格式化namenode和failovercontroler

failovercontroler也需要格式化: bin/hdfs zkfc -formatZK

bin/hdfs namenode -format -clusterid mycluster

在hadoop2节点上的namenode信息需要与hadoop1节点同步,不能通过简单的格式化做到,hadoop2节点上的namenode需要向hadoop1的namenode发送数据请求。因此我们还需要启动hadoop1上的namenode.

  在hadoop1上运行: bin/hdfs namenode

  在hadoop3上运行:bin/hdfs namenode  -bootstrapStandby

最后关闭hadoop1上的namenode,然后启动整个hadoop集群。

 

启动hadoop集群

cd /home/hadoop/hadoop-2.2.0

sbin/start-all.sh

从图上可以看出,先启动namenode,再启动datanode, 再启动ZK failover controller, 再启动resourcemanger, 最后启动nodemanager。

使用jps查看启动的进程

在hadoop1上运行jps

在hadoop2上运行jps

在hadoop3上运行jps

查看namenode的状态

  bin/hdfs haadmin -getServiceState hadoop1

  bin/hdfs haadmin -getServiceState hadoop2

从图上可以看出hadoop2上的namenode处于standby状态,而hadoop1上的namenode处于active状态

这些信息也可以通过Hadoop的web界面得到。

在浏览器里输入:http://hadoop1:50070

在浏览器里输入:http://hadoop2:50070

Failover 测试

从图上我们可以看出hadoop1节点上的namenode处于active状态,hadoop2上的节点处于standby状态,我们现在杀死hadoop1节点上的namenode,然后看hadoop2上的节点会自动变为active状态

在hadoop1上使用jps查看启动的进程

找到NameNode的进程,然后杀死它

  Kill -9 11146,发现namenode消失了

查看一下hadoop2节点的状态  bin/hdfs haadmin -getServiceState hadoop2

查看hadoop1节点的状态 bin/hdfs haadmin -getServiceState hadoop1

启动hadoop1节点上的namenode bin/hdfs namenode后

再查看hadoop1节点的状态 bin/hdfs haadmin -getServiceState hadoop1

很显然,hadoop1节点上namenode为standby状态,hadoop已经很好的解决了single namenode的问题,在不停机的条件下 备用节点成功的接管了主节点的任务。

尽管namenode可以很好的完成failover工作,但是他们之间使用nfs来存储变量的数据。nfs也会存在单点问题,也有可能停机导致整个集群的失败。Hadoop还提供了一种叫做jornalnode的技术,解决nfs的问题。

posted on 2014-05-13 09:56  cloudkiller  阅读(1279)  评论(0编辑  收藏  举报

导航