Christopher Olston from Yahoo Research presented Pig at the UCB DBMS lunch talk today. It's a framework that translates a high level processing specification into Hadoop jobs. Users only worry about providing functions for specific data parsing and computation. Pig and Hadoop do all the rest.

Here is a sample program (taken from slide 11) that joins two big datasets (Visits and Pages) to find sessions that end with the "best" page:

        Visits = load '/data/visits' as (user, url, time);
        Visits = foreach Visits generate user, Canonicalize(url), time;

         Pages = load '/data/pages' as (url, pagerank);

            VP = join Visits by url, Pages by url;
    UserVisits = group VP by user;
      Sessions = foreach UserVisits generate FindSessions(*);
  HappyEndings = filter Sessions by BestIsLast(*);

         store HappyEndings into '/data/happy_endings';

Pig is open-source. It's being "incubated" as an Apache project. More details here:

Apache Page: http://incubator.apache.org/pig

Subversion Repository: http://svn.apache.org/repos/asf/incubator/pig

Powerpoint slides: http://www.cs.cmu.edu/~olston/pig.ppt