运行Hadoop Pi示例

Question

I need some help running Hadoop Pi example in Windows. I have sucecssfully setup Hadoop on Windows and can see all NameNode, SecondaryNameNode, JobTracker and TaskTracker running. I can also hit urls:

http://localhost:50070/, http://localhost:50030 and http://localhost:50090/.

However when I am trying to run the in-built Pi example I am getting following exception.

$ hadoop jar hadoop/hadoop-examples-1.2.1.jar pi 4 100
Not a valid JAR: C:\cygwin\home\chandeln\hadoop\hadoop-examples-1.2.1.jar

I have also set both hadoop and java in my path.

$ java -version
java version "1.7.0_40"
Java(TM) SE Runtime Environment (build 1.7.0_40-b43)
Java HotSpot(TM) 64-Bit Server VM (build 24.0-b56, mixed mode)

$ hadoop version
Hadoop 1.2.1
Subversion https://svn.apache.org/repos/asf/hadoop/common/branches/branch-1.2 -r 1503152
Compiled by mattf on Mon Jul 22 15:23:09 PDT 2013
From source with checksum 6923c86528809c4e7e6f493b6b413a9a
This command was run using /C:/cygwin/usr/local/hadoop/hadoop-core-1.2.1.jar

My .bashrc looks like this:

export JAVA_HOME=/cygdrive/c/Java/jdk1.7.0_40
export HADOOP_PREFIX=/cygdrive/c/usr/local/hadoop
export PATH=$PATH:$HADOOP_PREFIX/bin

Can anybody shed some light on this ?

Answer 1

Even though Hadoop is written in Java, it does not play well on Windows. If you really want to get up and running with Hadoop in the shortest possible time, I would recommend using Linux. You can download the free VMWare Player application, and install it on Windows. Then, download the latest version on Ubuntu , and install it as a virtual machine using VMWare Player. Then you can use your Ubuntu virtual machine to install a single node cluster of Hadoop .

If this seems like too much work, open an Amazon AWS account, and find pre-configured Hadoop machines to work with.

Answer 2

That is most likely due to your Java version not matching with the Jar. Try with jdk1.6.

(Looks like you are playing with YDN Hadoop?)

Answer 3

尝试$ yarn jar hadoop / hadoop-examples-1.2.1.jar pi 4 100