Hewlett Packard Enterprise recently announced the release of the next version of Vertica: Frontloader (version 8). This new release includes advanced in-database machine learning as well as deployment support for Microsoft Azure.
Here’s a rundown of new features and changes in Vertica 8:
- Advanced Analytics / Machine Learning Package is now included in the server install. This package includes several machine learning functions such as sampling, k-means and logistical regression modeling. Geospatial Analytics are also included int he installer and spatial data can now be transformed between SRID (spatial reference identifiers).
- Directed Queries no support queries which invoke analytic functions as well as user-defined extensions.
- Multiple Constraint Violations will now be reported before the rollback of a SQL statement.
- Load support has been extended for the AUTO, DIRECT, and TRICKLE methods to two levels: table metadata and DML operations.
- Apache Spark is now supported which allows for the transfer of Vertica Tables to Spark DataFrames and vice versa.
- Management Console improvements include the ability to view history for Kafka and S3 jobs. Additional enhancements on charts and visualization for table utilization and extended monitoring.
- Several Apache Kafka functions have been updated and new features added.
- Vertica can now be deployed on Microsoft Azure through the Azure Marketplace.
Read more in the official press release.