Hadoop automatically moves your MapReduce program to where your data is so I think you can just do getHostName() (if you're using Java that is) and it should return the name of the node on which your program is running.
If you're using other languages such as Python, Ruby, etc. (i.e. using HadoopStreaming), the same idea holds true so you should be able to use the appropriate function/method available in those languages to get the host name.
0
Chris White
On
The configuration value fs.default.name will most probably give you a URL to the namenode, and if you get an instance of the FileSystem (Filesystem.get(conf)) you should be able to call the getUri() method to get the same information.
Hadoop automatically moves your MapReduce program to where your data is so I think you can just do getHostName() (if you're using Java that is) and it should return the name of the node on which your program is running.
If you're using other languages such as Python, Ruby, etc. (i.e. using HadoopStreaming), the same idea holds true so you should be able to use the appropriate function/method available in those languages to get the host name.