New in Cloudera Enterprise 5.11: Hue Data Search and Tagging

Categories: CDH Hadoop Hue

Self-service business intelligence and exploratory analytics continue to be a primary use case for Cloudera’s customers. Over the past year, we have made a number of significant advancements in Hue, the intelligent SQL editor, to provide a more powerful user experience for SQL developers and make them even more productive for those use cases.

The recent release of  Cloudera 5.11 furthers this effort with new enhancements around embedded search and tagging for faster data discovery, and SQL improvements for added efficiency. Read on to learn more about these enhancements and try out the new and improved Hue with one-click at demo.gethue.com.

Embedded Search & Tagging

Have you ever struggled to remember table names related to your project? Does it take much too long to find those columns or views? Hue now lets you easily search for any table, view, or column across all databases in the cluster. With the ability to search across tens of thousands of tables, you’re able to quickly find the tables that are relevant for your needs for faster data discovery.

In addition, you can also now tag objects with names to better categorize them and group them to different projects. These tags are searchable, expediting the exploration process through easier, more intuitive discovery.

Through an integration with Cloudera Navigator, existing tags and indexed objects show up automatically in Hue, any additional tags you add appear back in Cloudera Navigator, and the familiar Cloudera Navigator search syntax is supported.

To learn more, watch this video:

Hadoop Tutorial: Integrating Cloudera Navigator in Hue from The Hue Team on Vimeo.

To enable the Cloudera Navigator integration in Hue, login to Cloudera Manager, go to Hue > Configuration, and check the boxes for two Cloudera Navigator properties: Enable Navigator Metadata Server Integration and Enable Audit Collection.

For more detailed instruction, please refer to the documentation.

Enabling the feature in Cloudera Manager

Enabling the feature in Cloudera Manager

Usage

In Hue’s SQL Editor, a top search bar now appears. The autocomplete offers a list of facets and prefills the top values. Pressing enter lists the available objects, which can be opened and explored further in the sample popup, the assist or directly into the table browser app.

Granular Search

By default, only tables and views are returned. To search for columns, partitions, databases use the ‘type:’ filter.

Example of searches:

  • table:customer → Find the customer table
  • table:tax* tags:finance → List all the tables starting with tax and tagged with ‘finance’
  • owner:admin type:field usage → List all the fields created by the admin user that matches the usage string

Security

In secured clusters, where Sentry is configured, Hue makes sure that results only contain objects the user has access to according to its Sentry permissions. It also means that the facet search is simplified with no counts of the popular values.

Autocomplete listing the currently matching tables

Autocomplete listing the currently matching tables

Search results with preview and tag edition

Search results with preview and tag edition

SQL Improvements

The editor keeps getting better. Below you can read about a selection of recent major improvements.

Autocomplete popup revamp

Like in a regular code editor, the autocomplete is split in two parts. This is to prevent hangs while fetching the metadata of the current table or columns and to display more context information like the comments, type of objects, and their full names.

Autocomplete before

Autocomplete before

Autocomplete in new version

Autocomplete in new version

Apache Kudu primary keys are directly represented

Apache Kudu primary keys are directly represented

Create table wizard

If you’ve ever struggled with creating new SQL tables from files, you’ll be happy to learn that this is now much easier. With the latest Hue release, you can now create these in an ad hoc way and thereby expedite self-service analytics. The wizard has been revamped to two simple steps and also offers more formats. Now users just need to:

  1. Select a file
  2. Select the type of table

And that’s it! Files can be dragged & dropped, selected from HDFS or S3 (if configured), and their formats are automatically detected. The wizard also assists when performing advanced functionalities like table partitioning, Kudu tables, and nested types.

To learn more, watch this video:

Hadoop Tutorial – Create a table in Hue from The Hue Team on Vimeo.

We hope that the new version of Hue makes self-service data discovery and analytics easier and faster. If you have any questions or feedback, feel free to comment here, on the community forum, or via @gethue!

 

Facebooktwittergoogle_pluslinkedinmailFacebooktwittergoogle_pluslinkedinmail

2 responses on “New in Cloudera Enterprise 5.11: Hue Data Search and Tagging

  1. wayne zimmerman

    Navigator integration is a great idea – tags are OK – what would really be nice (and frankly much more useful than tags) is displaying (and managing the update of) the Navigator description property from within HUE (so it function as a data dictionary). Tags are really only half of the solution here (and not very useful by themselves)…. adding more properties from Navigator (like description, etc.) would add some real power and usability.

Leave a Reply

Your email address will not be published. Required fields are marked *

Prove you're human! *