Today we shipped Dremio 1.4!
This is a major release that includes many new features - see the highlights and more details below. Our products team will host a live discussion of these new features - you can register here.
- Enhanced vectorization performance. Dremio’s in-memory query execution engine is based on columnar data structures and formats provided by Apache Arrow. We improved the Java implementation of Arrow, and incorporated these enhancements into this release. You can read more about the changes here and the performance benefits here.
- Azure Data Lake Store Support. Azure Data Lake Store can now be used as a source and distributed storage option in Dremio.
- Enhanced query push downs for Elasticsearch. Dremio can now push down the NOT LIKE operator into Elasticsearch. This is also supported when Boolean expressions are part of a larger complex expression. In addition, parts of a predicate conjunction that can be pushed down will be pushed down, with the remaining components handled in the Dremio execution engine.
- Improved auto failover resiliency. Dremio now handles detecting master failures and electing a new master node more robustly.
- Better handling of non UTF-8 characters. Dremio now supports ISUTF8() and CONVERT_FROM() functions for dealing with non UTF-8 string values. Users can programmatically validate and replace/omit invalid characters as needed.
- Enhanced sampling for MongoDB. Dremio now uses the $sample stage for MongoDB versions 3.2 and later when previewing data from a collection.
- Learn more in the release notes here.
You can download here.
Be sure to follow the upgrade instructions for your deployment method.
If you have questions, you know where to ask.