Hadoop / HBase

http://photos1.meetupstatic.com/photos/event/1/e/f/2/600_268507922.jpeg

HBase is a non-relational distributed database inspired from Google BigTable.

It is a top-level Apache project, runs on top of Hadoop distributed file system, with features including:

• fault-tolerant storage of large quantities of sparse data

• data compression

• in-memory operations

• block cache and Bloom Filters on a per-column basis for real-time queries

• automatic sharding and failover

• query predicate push down

and many more..

Nicolas Liochon, an HBase committer and pmc member will detail what HBase is about, including some more specific aspects like transactional operations.

Contents in English, talk in French.