Hillsborough, CA

Oracle has been hinting at taking BerkelyDB aka Sleepycat into the NoSQL world for quite a few months.

Using Oracle Berkeley DB as a NoSQL Data Store

Today they made a formal announcement...along with many others. Not a lot has changed so far - just the name.

Lots has changed! It's an entire layer on top of the old JE edition that really makes it practically shared nothing and highly scalable. No JSON/REST, No C interface either...but it's a start.

Oracles features below with my [comments] afterwards.

Key-value pair data structure, keys are composed of Major & Minor keys [Example given was Customer ID as Major key while First Name and Last Name would be Minor keys]

Easy-to-use Java API with simple Put, Delete and Get operations [No C or JSON/REST yet]
Scalability
- Automatic, hash-function based data partitioning and distribution [seems well though out]
- Intelligent NoSQL Database driver is topology and latency aware, providing optimal data access [ immature - not Data Center or rack aware - can do stupid things easily right now.]
Predictable behavior
- ACID transactions, configurable globally and per operation
- Bounded latency via B-tree caching and efficient query dispatching [I guess it's completely bounded since no joins]
High Availability
- No single point of failure
- Built-in, configurable replication
- Resilient to single and multi-storage node failure
- Disaster recovery via data center replication
Easy Administration
- Web console or command line interface
- System and node management
- Shows system topology, status, current load, trailing and average latency, events and alerts