11.0.0 (Dremio November 2020) Release Notes

What's New

Custom Projects in AWS Editon

Dremio AWS Edition allows Dremio administrators to create custom projects that include customer-provided EBS volumes, EFS, and S3 buckets.

Preview: Vectorized Reader for Parquet-formatted Datasets

Dremio provides a vectorized reader for simple and complex types in Parquet-formatted datasets. Contact Dremio Customer Support to enable this feature.

Preview: Multiple Pushdown Filters for Parquet-formatted Datasets

Dremio supports pushing down multiple filters into Parquet-formatted datasets. Contact Dremio Customer Support to enable this feature.

Release Notes for Dremio JDBC Driver

Beginning with version 11.0.0, Dremio now provides release notes for the Dremio JDBC driver.

Other Enhancements

  • Dremio supports pushing down IN filters in Parquet-formatted datasets.
  • Dremio JDBC driver 11.0.0 supports TLS SNI.
  • Clarified error message regarding mismatched version tags when attempting to delete a Dremio object with the REST API.
  • Improved error messages for dremio-admin CLI tool.
  • 11.0.0 supports the following new functionality for Gandiva-based execution:
    • The Gandiva round() function now supports float and double data types.
    • Gandiva now supports the binary_string and split_part functions.

Deprecations

Dremio 11.0.0 deprecates legacy mode for relational and ARP-based (Advanced Relational Pushdown) external data sources.

Fixed Issues

Dremio doesn't read new files inserted into an external reflection until either restarting Dremio or the coordinator node
Fixed by invalidating the external reflection cache if tableRel has new data.

The Gandiva round() function doesn't support passing a single argument
Fixed.

The Upload query profiles to Dremio Support toggles in the Support Access pane of the Support page do nothing
Resolved by removing the toggles from the administrative user interface.

Unable to access ADLS Gen2 datasources after upgrading to Dremio 4.8.1
Fixed by reverting path access check on Azure storage plugin.

Restore command for admin-cli version 4.7.2 fails if data exceeds SQL column length of 32k bytes
Resolved by truncating the sql to 30k bytes while writing the Lucene index.


results matching ""

    No results matching ""