The Apache Phoenix team is pleased to announce its first releases as a top level project. Phoenix is the best performing open source SQL database for Apache HBase, a NoSQL data store. Phoenix is accessed as a JDBC driver and enables querying, updating, and managing HBase tables using SQL.
Our 3.1 release (for HBase 0.94.4+) includes:
- ~75 bug fixes
- Use of nested/derived tables in queries
- Automatic partial index rebuild on write failure
- Pig loader leveraging Phoenix SQL performance
- Easier usage of Phoenix with third-party SQL tools
- Secure cluster connection via URL
- Query over ANY/ALL for ARRAY type
- Specify CYCLE, MINVALUE, and MAXVALUE in sequence creation
Our 4.1 release (for HBase 0.98.1+) includes all of the above plus:
- Tracing leveraging Cloudera’s HTrace library to show per-request performance metrics all they way from the client through into the HBase server, and back again
- Local Indexes targeting write-heavy, space-constrained use cases by co-locating index and table data on the same region server
Source and binary downloads are available here.