The scripts install Pig on the master node so you can issue queries in Pig Latin from the master node. Here are a few pointers to get started with using it.

To run Pig, just type pig. This will start the Grunt shell, connected to the cluster. The following runs a simple Pig program against the sample data we loaded into HDFS above:

grunt> A = LOAD 'input';
grunt> B = FILTER A BY $0 MATCHES '.*dfs[a-z.]+.*';
grunt> DUMP B;
grunt> quit