Release notes

Imply 2.3.9 includes the following packages:

  • Druid 0.10.1-iap6
  • ImplyUI 0.3.33
  • PlyQL 0.11.2
  • Tranquility 0.8.2
  • ZooKeeper 3.4.10

Pivot evaluation

The Imply download includes a 30 day trial evaluation of the Imply interface (which includes Pivot, a powerful visual data explorer). Interface licenses are included with Imply subscriptions — contact us to learn more!

Upgrading from previous releases

When upgrading from early Imply 2.x releases, which bundled Druid 0.9.2, review the "Upgrading from 0.9.2" section of the release notes on https://github.com/druid-io/druid/issues/3944.

For notes about upgrading from pre-2.x versions, see the release notes for Imply 2.0.0.

Druid changes

This Imply release is based on Druid 0.10.1. Release notes for Druid 0.10.1 can be found at: https://github.com/druid-io/druid/releases/tag/druid-0.10.1. In addition, Imply 2.3.9 contains the following Druid patches:

  • TimestampSpec: Have "auto" detect timestamps in almost-iso format.
  • Kafka: Fixes needlessly low interpretation of maxRowsInMemory.
  • Get s3 objects directly from prefixes when listing is failed due to permission.
  • Implement get methods for PrefetchableFirehose.
  • free() dictionary merging buffers in IndexMerger.
  • Add metrics to the native queries underpinning SQL.
  • More informative QueryInterruptedException toString().
  • Replace Guava Enum.getIfPresent with builtin version.
  • Fix improper handling of empty arrays in StringDimensionIndexer.
  • Fix binary serialization in caching.
  • Fix GroupBy limit push down error when buffer is too small.
  • Fix decimal type handling in ORC extension.
  • Fix improper rounding of ORC decimals.
  • Return null for the null list in OrcStruct.
  • Fix race with SQL metadata caching.
  • Additional docs/log for direct memory usage.
  • Reduce excessive logging on the coordinator.

Imply interface (Pivot) changes

In this version Pivot was incorporated into the larger ImplyUI project the pivot executable and process is now imply-ui.

  • Minimum required node version changed to 4.5.0 (from 4.0.0)
  • Integrated SQL UI for running DruidSQL queries
  • Entirely new way of storing state, state storage is now mandatory and much more flexible
  • Added support for SQLite and Postgres state storage
  • Redesigned Pivot home screen with favourites
  • Added Grid table visualization
  • "Collections" renamed to "Dashboards", much more flexible layout system
  • All dashboards are now always filterable
  • Added ability to apply themes to data cubes and dashboards
  • Improvements to filtering on dashboards with tiles from different data cubes
  • Visualizations now look better at a small scale, multi-measure visualizations are supported in dashboards
  • Sharing and Access control lists for data cubes and dashboards
  • Short stateful URLs
  • Integrated LDAP connector
  • Added button to easily go back to the dashboard that you came form when expanding a tile
  • Ability to define a server side data cache to boost performance
  • Misc big fixes and improvements

Changes since 2.3.0

  • Added a new user type: dashboard-only-user which is a user that can only edit dashboards (not data cubes)
  • Added a new userMode: all-dashboard-only-users which will assign everyone as dashboard-only-user
  • Fixed bug in /mkurl
  • Only super-admin users can now edit the clusters in full imply UI mode
  • Ability to define altDataSource through data cube options
  • Fixes in over-aggressive caching
  • Fixes in shift+enter shortcut within the SQL UI
Overview