Jethro

Jethro is a SQL-on-Hadoop query and business intelligence acceleration layer for interactive analytics on large datasets. It sits between Hadoop-compatible storage or Amazon S3 and BI tools such as Tableau, Qlik, and MicroStrategy, allowing users to submit SQL queries through JDBC or ODBC.

Jethro loads data into a query-optimized column and index structure stored in HDFS or Amazon S3. It uses automatically generated, compressed bitmap indexes, caching, and automatically created micro-cubes to reduce the amount of data read for a query. This approach is intended to support interactive filtering, drill-down, aggregation, and dashboard workloads against multi-billion-row datasets without requiring every query to scan the full source dataset.

The platform uses dedicated query hosts for SQL processing while leaving the underlying Hadoop cluster available for storage and other compute workloads. Query nodes are stateless and can be added to increase concurrency, while the loader service ingests source data and creates the associated index and column files. Jethro supports deployments alongside common Hadoop distributions and can use Hadoop or S3 as its storage layer.

Jethro is designed for organizations that use Hadoop-based or object-storage data platforms but need faster SQL response times for BI dashboards and exploratory analysis than their existing query engine can provide.