Hadoop集群配置

一、系统文件配置

1. 集群部署规划

• NameNode和SecondaryNameNode不要安装在同一台服务器

• ResourceManager也很消耗内存，不要和NameNode、SecondaryNameNode放在同一台机器上。

这里装了四台机器，ant151,ant152,ant153,ant154。

ant151	ant152	ant153	ant154
NameNode	NameNode
DataNode	DataNode	DataNode	DataNode
NodeManager	NodeManager	NodeManager	NodeManager
		ResourceManager	ResourceManager
JournalNode	JournalNode	JournalNode
DFSZKFController	DFSZKFController
zk0	zk1	zk2

2. 配置文件说明

Hadoop配置文件分为默认配置文件和自定义配置文件，只有用户想修改某一默认配置值时，才需要修改自定义配置文件。

core-site.xml、hdfs-site.xml、mapred-site.xml、yarn-site.xml四个配置文件放在$HADOOP_HOME/etc/hadoop路径下。

3.配置集群

core-site.xml

    <property><name>fs.defaultFS</name><value>hdfs://gky</value><description>逻辑名称，必须与hdfs-site.xml中的dfs.nameservices值保持一致</description></property><property><name>hadoop.tmp.dir</name><value>/opt/soft/hadoop313/tmpdata</value><description>namenode上本地的hadoop临时文件夹</description></property><property><name>hadoop.http.staticuser.user</name><value>root</value><description>默认用户</description></property><property><name>hadoop.proxyuser.root.hosts</name><value>*</value><description></description></property><property><name>hadoop.proxyuser.root.groups</name><value>*</value><description></description></property><property><name>io.file.buffer.size</name><value>131072</value><description>读写文件的buffer大小为：128K</description></property><property><name>ha.zookeeper.quorum</name><value>ant151:2181,ant152:2181,ant153:2181</value><description></description></property><property><name>ha.zookeeper.session-timeout.ms</name><value>10000</value><description>hadoop链接zookeeper的超时时长设置为10s</description></property>

hdfs-site.xml

    <property><name>dfs.replication</name><value>3</value><description>Hadoop中每一个block的备份数</description></property><property><name>dfs.namenode.name.dir</name><value>/opt/soft/hadoop313/data/dfs/name</value><description>namenode上存储hdfs名字空间元数据目录</description></property><property><name>dfs.datanode.data.dir</name><value>/opt/soft/hadoop313/data/dfs/data</value><description>datanode上数据块的物理存储位置</description></property><property><name>dfs.namenode.secondary.http-address</name><value>ant151:9869</value><description></description></property><property><name>dfs.nameservices</name><value>gky</value><description>指定hdfs的nameservice,需要和core-site.xml中保持一致</description></property><property><name>dfs.ha.namenodes.gky</name><value>nn1,nn2</value><description>gky为集群的逻辑名称，映射两个namenode逻辑名</description></property><property><name>dfs.namenode.rpc-address.gky.nn1</name><value>ant151:9000</value><description>namenode1的RPC通信地址</description></property><property><name>dfs.namenode.http-address.gky.nn1</name><value>ant151:9870</value><description>namenode1的http通信地址</description></property><property><name>dfs.namenode.rpc-address.gky.nn2</name><value>ant152:9000</value><description>namenode2的RPC通信地址</description></property><property><name>dfs.namenode.http-address.gky.nn2</name><value>ant152:9870</value><description>namenode2的http通信地址</description></property><property><name>dfs.namenode.shared.edits.dir</name><value>qjournal://ant151:8485;ant152:8485;ant153:8485/gky</value><description>指定NameNode的edits元数据的共享存储位置(JournalNode列表)</description></property><property><name>dfs.journalnode.edits.dir</name><value>/opt/soft/hadoop313/data/journaldata</value><description>指定JournalNode在本地磁盘存放数据的位置</description></property>    <!-- 容错 --><property><name>dfs.ha.automatic-failover.enabled</name><value>true</value><description>开启NameNode故障自动切换</description></property><property><name>dfs.client.failover.proxy.provider.gky</name><value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value><description>失败后自动切换的实现方式</description></property><property><name>dfs.ha.fencing.methods</name><value>sshfence</value><description>防止脑裂的处理</description></property><property><name>dfs.ha.fencing.ssh.private-key-files</name><value>/root/.ssh/id_rsa</value><description>使用sshfence隔离机制时，需要ssh免密登陆</description></property>    <property><name>dfs.permissions.enabled</name><value>false</value><description>关闭HDFS操作权限验证</description></property><property><name>dfs.image.transfer.bandwidthPerSec</name><value>1048576</value><description></description></property>    <property><name>dfs.block.scanner.volume.bytes.per.second</name><value>1048576</value><description></description></property>

mapred-site.xml

    <property><name>mapreduce.framework.name</name><value>yarn</value><description>job执行框架： local, classic or yarn</description><final>true</final></property><property><name>mapreduce.application.classpath</name><value>/opt/soft/hadoop313/etc/hadoop:/opt/soft/hadoop313/share/hadoop/common/lib/*:/opt/soft/hadoop313/share/hadoop/common/*:/opt/soft/hadoop313/share/hadoop/hdfs/*:/opt/soft/hadoop313/share/hadoop/hdfs/lib/*:/opt/soft/hadoop313/share/hadoop/mapreduce/*:/opt/soft/hadoop313/share/hadoop/mapreduce/lib/*:/opt/soft/hadoop313/share/hadoop/yarn/*:/opt/soft/hadoop313/share/hadoop/yarn/lib/*</value></property><property><name>mapreduce.jobhistory.address</name><value>ant151:10020</value></property><property><name>mapreduce.jobhistory.webapp.address</name><value>ant151:19888</value></property><property><name>mapreduce.map.memory.mb</name><value>1024</value><description>map阶段的task工作内存</description></property><property><name>mapreduce.reduce.memory.mb</name><value>2048</value><description>reduce阶段的task工作内存</description></property>

yarn-site.xml

    <property><name>yarn.resourcemanager.ha.enabled</name><value>true</value><description>开启resourcemanager高可用</description></property><property><name>yarn.resourcemanager.cluster-id</name><value>yrcabc</value><description>指定yarn集群中的id</description></property><property><name>yarn.resourcemanager.ha.rm-ids</name><value>rm1,rm2</value><description>指定resourcemanager的名字</description></property><property><name>yarn.resourcemanager.hostname.rm1</name><value>ant153</value><description>设置rm1的名字</description></property><property><name>yarn.resourcemanager.hostname.rm2</name><value>ant154</value><description>设置rm2的名字</description></property><property><name>yarn.resourcemanager.webapp.address.rm1</name><value>ant153:8088</value><description></description></property><property><name>yarn.resourcemanager.webapp.address.rm2</name><value>ant154:8088</value><description></description></property>    <property><name>yarn.resourcemanager.zk-address</name><value>ant151:2181,ant152:2181,ant153:2181</value><description>指定zk集群地址</description></property><property><name>yarn.nodemanager.aux-services</name><value>mapreduce_shuffle</value><description>运行mapreduce程序必须配置的附属服务</description></property><property><name>yarn.nodemanager.local-dirs</name><value>/opt/soft/hadoop313/tmpdata/yarn/local</value><description>nodemanager本地存储目录</description></property><property><name>yarn.nodemanager.log-dirs</name><value>/opt/soft/hadoop313/tmpdata/yarn/log</value><description>nodemanager本地日志目录</description></property><property><name>yarn.nodemanager.resource.memory-mb</name><value>2048</value><description>resource进程的工作内存</description></property><property><name>yarn.nodemanager.resource.cpu-vcores</name><value>2</value><description>resource工作中所能使用机器的内核数</description></property><property><name>yarn.scheduler.minimum-allocation-mb</name><value>256</value><description></description></property><property><name>yarn.log-aggregation-enable</name><value>true</value><description></description></property><property><name>yarn.log-aggregation.retain-seconds</name><value>86400</value><description>日志保留多少秒</description></property><property><name>yarn.nodemanager.vmem-check-enabled</name><value>false</value><description></description></property><property><name>yarn.application.classpath</name><value>/opt/soft/hadoop313/etc/hadoop:/opt/soft/hadoop313/share/hadoop/common/lib/*:/opt/soft/hadoop313/share/hadoop/common/*:/opt/soft/hadoop313/share/hadoop/hdfs/*:/opt/soft/hadoop313/share/hadoop/hdfs/lib/*:/opt/soft/hadoop313/share/hadoop/mapreduce/*:/opt/soft/hadoop313/share/hadoop/mapreduce/lib/*:/opt/soft/hadoop313/share/hadoop/yarn/*:/opt/soft/hadoop313/share/hadoop/yarn/lib/*</value><description></description></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><description></description></property>

hadoop-env.sh

export JAVA_HOME=/opt/soft/jdk180
export HDFS_NAMENODE_USER=root
export HDFS_DATANODE_USER=root
export HDFS_SECONDARYNAMENODE_USER=root
export HDFS_JOURNALNODE_USER=root
export HDFS_ZKFC_USER=root
export YARN_RESOURCEMANAGER_USER=root
export YARN_NODEMANAGER_USER=root

workers

ant151
ant152
ant153
ant154

二、集群首次启动

1. 启动zk集群

可以直接运行脚本文件

代码：

[root@ant151 shell]# ./zkop.sh start 

2. 启动ant151,ant152,ant153的journalnode服务：

[root@ant151 shell]# hdfs --daemon start journalnode

3. 在ant151格式化hfds namenode:

[root@ant151 shell]# hdfs namenode -format

4. 在ant151启动namenode服务：hdfs --daemon start namenode

[root@ant151 shell]# hdfs --daemon start namenode

5. 在ant152机器上同步namenode信息

[root@ant151 shell]# hdfs namenode -bootstrapStandby

6. 在ant152启动namenode服务：hdfs --daemon start namenode

[root@ant152 soft]# hdfs --daemon start namenode

查看namenode节点状态：hdfs haadmin -getServiceState nn1|nn2

[root@ant152 soft]# hdfs haadmin -getServiceState nn1

7. 关闭所有dfs有关的服务

[root@ant151 soft]# stop-dfs.sh

8. 格式化zk

[root@ant151 soft]# hdfs zkfc -formatZK

9. 启动dfs

[root@ant151 soft]# start-dfs.sh

10. 启动yarn: [root@ant151 soft]# start-yarn.sh

[root@ant151 soft]# start-yarn.sh

11. 查看resourcemanager节点状态

[root@ant151 soft]# yarn rmadmin -getServiceState rm1

rm1状态：standby

rm2状态：active

当前进程状态：

kill掉active进程

尝试访问，无法链接

恢复ant152的namenode进程