INSPIRE has migrated from development to the production D4Science's Hadoop facility. A large collection of data (about 200 thousand papers) was uploaded to the Hadoop Distributed File System pool. After running initial tests, the cluster was configured and tuned for scalable processing which is now being performed.