One
Eclipse Import Hadoop Source project
Basic steps:
1) Create a new Java project "hadoop-1.2.1" in Eclipse
2) Copy the Core,hdfs,mapred,tools,example four directory under the directory src of the Hadoop compression package to the SRC directory of the new project above
3) Right click to select Build path, modify Java Build path "source", delete src, add src/core,src/hdfs,src/mapred,src/tools several source directories;
4) Create a new Lib directory in the project, add the project dependent jar package, import all the jar packages in the Hadoop decompression directory (remove two documents), do not miss the jar package in its subdirectory jsp-2.1, import all the jar packages under the Ant program Lib
5) Select all the jar packages in the Lib directory, then right-click to select Build path to add
6) Right click on Project selection build project, recompile project, find error
7) Import the ant package, import all the jar packages under ant into the Lib directory, right-click the project selection build Path-->libraries-->add Jars-->hadoop-1.2.1-->lib will appear as the added package , click OK to finish.
8) in compiling the project, found or wrong, then right click on Project selection build path-->libraries find JRE System library--> Click access rules:no rules Defined- Select Edit
Resolution box changed to Accessible,rule pattern fill **/* dot ok
This completes.
Ii. writing Hadoop programs under eclipse
Basic steps:
1) New Hadoop project, in the project to create a new Lib directory, copy the Hadoop compression package extracted from the directory lib several jar packages, respectively:
Where Hadoop-core-1.2.1.jar is in the Hadoop-1.2.1 directory and then added to the build path path.
2) Add a configuration file for Hadoop
Click on the project to create a new folder, named Conf, you need to put two configuration files, respectively, Core-site.xml and Hdfs-site.xml
3) Add Unit Test Package JUNIT4
Select Project Right-click to select Build Path-->libraries-->add library-->junit-->junit4-->ok
In this case, we can set up the package and test class to add, delete, change and check the HDFs under the project SRC.
Example: Look at a file class volume under HDFs and display it on the console with the following code:
PackageOrg.chaofn.hadoop.hdfs;ImportJava.io.InputStream;ImportJava.net.URL;Importorg.apache.hadoop.fs.FsUrlStreamHandlerFactory;Importorg.apache.hadoop.io.IOUtils;Importorg.junit.Test; Public classHdfsurltest {//Let the Java program identify the URL of HDFs Static{url.seturlstreamhandlerfactory (Newfsurlstreamhandlerfactory ()); } //View File Contents@Test Public voidTestread ()throwsexception{InputStream in=NULL; //file pathString fileurl= "Hdfs://linux.chaofn.org:9000/wc/input/core-site.xml"; Try{ in=NewURL (FILEURL). OpenStream (); //read the contents of the file and print it to the consoleIoutils.copybytes (in, System.out, 4096,false); }finally{ioutils.closestream (in); } } }
Eclipse Imports Hadoop Source project and writes Hadoop program