[解决]Spark运行中java.net.UnknownHostException: nameservice1的异常

[Author]: kwu

解决Spark运行中java.net.UnknownHostException: nameservice1的异常，nameservice1是配置hdfs中用到的HA，在执行spark程序时出现 java.net.UnknownHostException: nameservice1异常，找到正确的路径。

分析在CDH5.4中配置文件/etc/hadoop/conf/hdfs-site.xml，此文件记录了hdfs的路径与主节点的配置

hdfs-site.xml 主要配置如下：

[html] view
plaincopy

<property>

<name>dfs.nameservices</name>

<value>nameservice1</value>

</property>

<property>

<name>dfs.client.failover.proxy.provider.nameservice1</name>

<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>

</property>

<property>

<name>dfs.ha.automatic-failover.enabled.nameservice1</name>

<value>true</value>

</property>

<property>

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

<value>bdc40.hexun.com:2181,bdc41.hexun.com:2181,bdc46.hexun.com:2181,bdc53.hexun.com:2181,bdc54.hexun.com:2181</value>

</property>

<property>

<name>dfs.ha.namenodes.nameservice1</name>

<value>namenode50,namenode85</value>

</property>

<property>

<name>dfs.namenode.rpc-address.nameservice1.namenode50</name>

<value>bdc20.hexun.com:8020</value>

</property>

<property>

<name>dfs.namenode.servicerpc-address.nameservice1.namenode50</name>

<value>bdc20.hexun.com:8022</value>

</property>

<property>

<name>dfs.namenode.http-address.nameservice1.namenode50</name>

<value>bdc20.hexun.com:50070</value>

</property>

<property>

<name>dfs.namenode.https-address.nameservice1.namenode50</name>

<value>bdc20.hexun.com:50470</value>

</property>

<property>

<name>dfs.namenode.rpc-address.nameservice1.namenode85</name>

<value>bdc220.hexun.com:8020</value>

</property>

<property>

<name>dfs.namenode.servicerpc-address.nameservice1.namenode85</name>

<value>bdc220.hexun.com:8022</value>

</property>

<property>

<name>dfs.namenode.http-address.nameservice1.namenode85</name>

<value>bdc220.hexun.com:50070</value>

</property>

<property>

<name>dfs.namenode.https-address.nameservice1.namenode85</name>

<value>bdc220.hexun.com:50470</value>

</property>

配置了 nameservice1的别名，对应两台namenode50,namenode85的节点上。

通过zookeeper来选举出当前的namenode的active.

hdfs-site.xml文件复制，到$SPARK_HOME/conf 目录下，即可解决spark job 访问hdfs路径的问题。