java - 如何定义HADOOP类路径？

Question

此代码来自 Hadoop 和 Spark 类(class)。我在 VM(HDP-sandbox)上运行它

[root@sandbox-hdp Lesson-5.3_hadoop-grep]# javac -classpath /usr/lib/hadoop/hadoop-core.jar -d Grep_classes Grep.java
Grep.java:22: error: package org.apache.hadoop.conf does not exist
import org.apache.hadoop.conf.Configuration;
                             ^
Grep.java:23: error: package org.apache.hadoop.conf does not exist
import org.apache.hadoop.conf.Configured;
                             ^
Grep.java:24: error: package org.apache.hadoop.fs does not exist
import org.apache.hadoop.fs.FileSystem;
                           ^
Grep.java:25: error: package org.apache.hadoop.fs does not exist
import org.apache.hadoop.fs.Path;

我是Hadoop新手，如何定义类路径？

如果我尝试

hadoop classpath

输出

/usr/hdp/3.0.1.0-187/hadoop/conf:/usr/hdp/3.0.1.0-187/hadoop/lib/*:/usr/hdp/3.0.1.0-187/hadoop/.//*:/usr/hdp/3.0.1.0-187/hadoop-hdfs/./:/usr/hdp/3.0.1.0-187/hadoop-hdfs/lib/*:/usr/hdp/3.0.1.0-187/hadoop-hdfs/.//*:/usr/hdp/3.0.1.0-187/hadoop-mapreduce/lib/*:/usr/hdp/3.0.1.0-187/hadoop-mapreduce/.//*:/usr/hdp/3.0.1.0-187/hadoop-yarn/./:/usr/hdp/3.0.1.0-187/hadoop-yarn/lib/*:/usr/hdp/3.0.1.0-187/hadoop-yarn/.//*:/usr/hdp/3.0.1.0-187/tez/*:/usr/hdp/3.0.1.0-187/tez/lib/*:/usr/hdp/3.0.1.0-187/tez/conf:/usr/hdp/3.0.1.0-187/tez/conf_llap:/usr/hdp/3.0.1.0-187/tez/doc:/usr/hdp/3.0.1.0-187/tez/hadoop-shim-0.9.1.3.0.1.0-187.jar:/usr/hdp/3.0.1.0-187/tez/hadoop-shim-2.8-0.9.1.3.0.1.0-187.jar

如何更改 javac 行？

Answer 1

您可以在hadoop-env.sh中导出HADOOP_CLASSPATH。您可以向 HADOOP_CLASSPATH 添加更多 jar。

稍后在运行作业之前在 shell 中运行以下命令。

export HADOOP_CLASSPATH=$(hadoop classpath)
echo $HADOOP_CLASSPATH

For reference see this