Big Data Service 3.0.29 has added additional features

To view the ODH version, JDK version, and OS version for this release, see Big Data Service Release and Patch Versions.

Big Data Service 3.0.29 has the following new features and updates:

Updated Hive support for two Apache Hive fixes:
- Fixed Vectorized PTF for count and distinct over partition failure.
- Fixed issue for drop table involved in materialized view leaves the table in inconsistent state.
Identity Domain integration enables OCI Identity Domain users to access Big Data Service 3.0.29 or later cluster resources and also Big Data Service 3.0.29 or later cluster users to access OCI services.
Change the node type shape when starting a cluster.
Add/remove NAT gateway after cluster creation.
Supports API key generation in Big Data Service for users belonging to any identity domain.
Performance tuning steps to optimize the performance and scalability of Trino for customer specific use cases.
Updated ODH patch installation strategy to include task failure tolerance per batch option so that the maximum number of possible failures don't exceed 20% of the nodes in the cluster.
Updated documentation to include information on enabling Hive HA and Oozie HA.
Manage Big Data Service schemas with APIs.

Bug fixes:

Resolved issues with node addition failures in ZooKeeper SSL-enabled clusters.
Added fix for Hue start up issue when multiple clusters are configured in HDFS nameservices.
Added fix for missing HBase client JAR as a part of Oozie classpath.
Fixed Oozie Error and Audit log rendering in Oozie UI.
Resolved logs rendering issues from Hue UI.
Resolved Unwanted Error and Warn logs from being displayed from console in Spark Shell,Pyspark and Spark-SQL.
Add SSL support for JupyterHub on Oracle Linux 8.
Fixed Schema Registry issue that caused prolonged delays during service shutdown and start up operations.
MySQL database password field made mandatory while adding services such as Hive in Ambari; Hive, Oozie, JupyterHub, Hue, and Schema Registry services. The database password is automatically updated in the backend MySQL database when you change Ambari configurations. You don't need to run update_mysql_password as a prerequisite.
Resolved Ambari alerts issue for the Trino service, ensuring notifications are triggered when the Trino Coordinator or Worker processes aren't running.
Livy server session timeout for Spark has been increased from 5 minutes to 30 minutes for ODH 2 clusters.
Fixed Hadoop Data node Web UI Ambari alert definition so that clusters with ODH 2.0.9 or later don't receive an alert for Hadoop Data node Web UI on SSL enabled clusters.