需要进一步学习hadoop、需要看看内部源码实现,因此需要将hadoop源码导入都eclipse中,简单总结一下,具体步骤如下:

首先确保已经安装了git、maven3、protobuf2.5、如果没有安装需要提前安装一下

1、下载hadoop源码

git clone git://git.apache.org/hadoop-common.git

2、进入hadoop-common目录,用maven生成eclipse工程

mvn install -DskipTests
mvn eclipse:eclipse -DdownloadSources=true -DdownloadJavadocs=true

3、导入Eclipse

导入Hadoop-Common

File->import...->Choose "Existing Projects into Workspace" 选择hadoop-common-project目录为跟目录导入

File->import...->Choose "Existing Projects into Workspace" 选择hadoop-assemblies目录为跟目录导入

导入HDFS

File->import...->Choose "Existing Projects into Workspace" 选择hadoop-hdfs-project目录为跟目录导入

导入MapReduce

File->import...->Choose "Existing Projects into Workspace" 选择hadoop-mapreduce-project目录为跟目录导入

导入YARN

File->import...->Choose "Existing Projects into Workspace" 选择hadoop-yarn-project目录为跟目录导入

OK、全部导入成功,但是需要注意两个地方:

首先如果想hadoop-common全部编译成功,需要将hadoop-common的target/generate-test-source/java目录设置source folder

其次还得把JDK lib目录的tools.jar添加到classpath中。

配置好上面之后,所有代码都清晰可见


转自:http://blog.csdn.net/yangfei001/article/details/36898929