mmadan: Nutch is a full web crawling system which uses Hadoop. It has been around for many years - and in fact could be credited with creating Hadoop.
I tried supporting Nutch for a while (Not through Hortonworks of course), but it is still very much R&D software because there are so few companies using it. There is some significant confusion about moving away from MapReduce to YARN.
Stanbol is something I am less familiar with - but since it consists of LOTS of Apache projects I think it would be as complicated as Hadoop to support.