Skip to main content
Version: 1.3.1.0

Known Issues — ODP 1.3.1.0

Overview​

This page documents known issues and limitations in ODP 1.3.1.0. Where a workaround is available, it is described below. Clemlab tracks these issues in the GitHub issue tracker and updates them as fixes become available.

Hue — Experimental Status​

Hue is included in ODP as an experimental component. It is not part of the standard ODP test matrix and may exhibit instability, especially in Kerberos-secured clusters or when integrated with multiple backend services simultaneously.

Hue is Experimental

Hue is provided for evaluation purposes only. It is not recommended for production deployments and does not receive the same level of support as GA components. Use the Ambari Views or service-native UIs (e.g., Ranger Admin UI, Atlas UI, Oozie Web Console) for production workloads.

Known limitations with Hue in ODP 1.3.1.0:

  • Hue may fail to authenticate against HiveServer2 in clusters with Kerberos + SSL enabled simultaneously
  • The Hue Spark SQL editor does not support Spark 3.4+ features (e.g., session catalog switching)
  • Hue's HDFS file browser may time out on directories with more than 10,000 entries

Polaris REST Catalog — Tech Preview​

The Polaris REST Catalog is a Tech Preview component in ODP 1.3.1.0.

Tech Preview

Polaris is not recommended for production use. The upgrade path between ODP versions for Polaris metadata is not guaranteed. See the Tech Preview Features page for details.

Known limitations:

  • Polaris does not support multi-realm configurations through Ambari in this release
  • Polaris catalog bootstrap must be performed manually after service start (Ambari does not automate initial catalog creation)
  • Integration with Ranger for fine-grained Polaris access control is not yet available

Kudu — Tech Preview​

Apache Kudu is a Tech Preview component in ODP 1.3.1.0.

Tech Preview

Kudu is not recommended for production use. See the Tech Preview Features page for details.

Known limitations:

  • Kudu tablet servers require SSD storage; performance on spinning disks is significantly degraded and not supported
  • Co-locating Kudu tablet servers with HDFS DataNodes on the same host is not recommended and may cause resource contention
  • Kudu replication defaults (factor 3) require a minimum of 3 worker nodes with Kudu tablet server roles assigned
  • The Ambari Kudu service definition does not yet support rolling restarts

Reporting New Issues​

If you encounter an issue not listed here, please report it at https://github.com/clemlabprojects with:

  • ODP version and component versions
  • Operating system and JDK version
  • Steps to reproduce
  • Relevant log files from /var/log/<component>/