Home

v4.5.2.2 Release notes

Focused on Amazon EMR, Azure HDInsight, Azure Databricks, and Google Cloud Dataproc.
Software version

Release Date: 10/22/2019

See v4.5.2.2 for download information.

Software Upgrade support
  • Upgrade from 4.5.0.5 and 4.5.2.x only. See here for specific upgrade information.

Unravel sensor upgrade
  • You must install the new sensor when upgrading from Unravel v4.5.0.5.

Certified platforms

Please review your platform's compatibility matrix below before you upgrade/install.

Updates to Unravel's configuration properties
  • None

Unsupported

All On-Demand Features

  • Sessions

  • Reports

    • Capacity Forecasting

    • Migration Planning

    • Cluster Optimization

    • File Reports

    • Queue Analysis

    • Small Files

    • Top X

  • AutoActions

  • Incorporating true “workflows”, e.g., Databricks jobs orchestration via services like ADF.

  • We have not examined the functionality provided by Cluster Pools and how it can affect Unravel functionality.

  • No support for "interactive clusters"; only Job clusters supported including when New Job Clusters are created.

  • Spark streaming applications have not been tried out or tested.

  • Unravel's APIs

  • Clusters created without Unravel Initialization actions and have auto-scaling setup are not supported. (DTP-9)

  • Clusters created without Unravel bootstrap and that have an auto-scaling setup are not supported. (EMR-83)

  • Automatic configuration when Kafka cluster scales up/down.

  • Linking Spark Streaming app topic to Kafka page.

  • ATSv2.0

  • Head node HA.

  • YARN aggregate logs are not captured for MR jobs. Therefore, the

    • Logs tab for MR jobs is empty.

    • Information in the Errors tab which is derived from the logs isn't available.

  • AutoActions

    • Kill and move actions. (AA-103)

    • Rules that span across multiple clusters, i.e., you can not specify a rule that aggregates metrics from multiple clusters and monitor its violation. (AA-131)

  • AMI/CFN/ARM Template based installation support.

  • No special support for cloud-native constructs, e.g., IAM roles for RBAC etc.

  • Multi-Metastore

    • Unsupported scenario

      When a cluster has more than one metastore or if an external metastore is used (as opposed to the one that comes default with the cluster).

    • Also, the Unravel node, at a given time, can retrieve data from one metastore (it needs to be configured as described in Hive Metastore configuration documentation).

  • Workflows

    • Airflow

  • Unravel for Google Cloud Dataproc is not available via GCP Marketplace yet.

New features and improvements
  • API no longer fails for Spark app with 1000+ stages (UIX-2054)

Bug fixes
  • API no longer fails for Spark app with 1000+ stages. (UIX-2054)

Known issues
  • gcLoad metrics shown under the Resource tab for Spark are incorrect. (CUSTOMER-1094)

For support issues, visit Unravel Support.