What’s New in this Release

New features and changes for DAS/DAS-Lite have been introduced in the 1.3.0 release, along with documentation updates.

• You can now authenticate the users by using the following methods:
  • Knox proxy
  • SPNEGO

  You can also choose not to set up user authentication. In this case, all the queries are executed as a hive service user.

  A new field, user_authentication has been introduced in the Ambari configuration where you can specify the method you want to use to authenticate your users.

• For added privacy, only the admin users can access the Read/Write report and the Join report. The non-admin users lack the privileges to view the Reports page or the Reports tab.
• DAS/DAS-Lite supports authorization with Ranger, if Ranger is configured for Hive. It uses the same access rules as configured for Hive. However, there is an exception for the admin users that are configured in DAS. The admin users that are configured in DAS will still be able to view all the databases, tables, and queries in DAS.
• The Tez DAG events are exposed in the dag_data table, and the Hive query events are exposed in the query_data table in the sys database. You can run Hive queries to process these events and generate custom reports. This feature is supported on the HDP 2.6.5 and HDP 3.1.0.31 releases.
• The DAG info swimlane shows detailed, task-level information for every vertex within the DAG flow. You can further view the task attempts for a particular task, and you can group the result set either by tasks, containers, or nodes. You can also download logs for a particular task attempt.
• DAS has a cleanup mechanism that, by default, purges all the queries and DAG information older than 30 days and purges old reports after 365 days from the DAS Postgres database to optimize its available capacity. You also can customize the cleanup frequency and time.
• DAS displays an alert if the Hive replication (repl dump) has not occurred for more than 15 minutes. The time when the replication last occurred is displayed in the Last Updated field on the Compose page and the Database page.
• Keyboard shortcut for executing a query: You can press Ctrl + Enter keys on a Windows system or control + return keys on an OSX system to start query execution from the query editor on the Compose page. This operation is a keyboard shortcut for the EXECUTE button.
• On the Join report, you can view all the joins with a particular column by moving the mouse pointer over a particular column name. This mouse action highlights all the connection lines originating from the selected column in a different color. You can also get a join count (number of times the two columns were joined during the selected time period) by moving the mouse pointer over the line connecting the two columns.
• The Join report also displays the database name to which a table belongs.