Support Changes
This section describes the support changes in version 10.2.2.
Hive Engine
Effective in version 10.2.2, Informatica dropped support for the Hive mode of execution for jobs run in the Hadoop environment. You cannot configure or run jobs on the Hive engine.
Informatica continues to support the Blaze and Spark engines in the Hadoop environment, and it added support for the Databricks Spark engine in the Databricks environment.
You need to update all mappings and profiles configured to run on the Hive engine before you upgrade.
For more information, see the Informatica 10.2.2 upgrade guides.
Distribution Support
Informatica big data products support Hadoop and Databricks environments. In each release, Informatica adds, defers, and drops support for Hadoop distribution versions. Informatica might reinstate support for deferred versions in a future release.
Big Data Management added support for the Databricks environment and supports the Databricks distribution version 5.1.
The following table lists the supported Hadoop distribution versions for Informatica 10.2.2 big data products:
Product | EMR | HDI | CDH | HDP | MapR |
|---|
Big Data Management | 5.16 | 3.6.x | 5.15 5.16 | 2.6.x | 6.0.x MEP 5.0 |
Enterprise Data Catalog | N/A | 3.6.x Note: HDI is supported only for WASB non-Kerberos. | 5.15 5.16 Note: You can use OpenJDK 1.8.0 only on Enterprise Data Catalog deployed on a CDH 5.16 Hadoop distribution. | 2.6.x | N/A |
Enterprise Data Preparation | 5.16 | 3.6.x Note: HDI is supported only for ADLS and WASB non-Kerberos. | 5.15 5.16 | 2.6.x | 6.0.x MEP 5.0 |
To see a list of the latest supported versions, see the Product Availability Matrix on the Informatica Customer Portal:
https://network.informatica.com/community/informatica-network/product-availability-matrices.
Support Changes for Big Data Management
This section describes the support changes to Big Data Management in version 10.2.2.
- Python Transformation
- Effective in version 10.2.2, support for binary ports in the Python transformation is deferred. Support will be reinstated in a future release.
Support Changes for Big Data Streaming
This section describes the changes to Big Data Streaming in version 10.2.2.
- Azure Event Hubs
- Effective in version 10.2.2, upgraded streaming mappings become invalid. You must re-create the physical data objects to run the mappings in Spark engine that uses Spark Structured Streaming. After you re-create the physical data objects, the following properties are not available for Azure Event Hubs data objects:
- - Consumer Properties
- - Partition Count
For more information, see the Informatica 10.2.2 Upgrade Guide.
- Data object types
- Effective in version 10.2.2, support for some data object types is deferred. Support will be reinstated in a future release.
The following table describes the deferred support for data object types in version 10.2.2:
Object Type | Object |
|---|
Source | JMS MapR Streams |
Target | MapR Streams |
Transformation | Data Masking Joiner (master outer, detail outer, and full outer join types) Rank Sorter |
- For more information, see the Informatica Big Data Streaming 10.2.2 User Guide.
Universal Connectivity Framework in Enterprise Data Catalog
Effective in version 10.2.2, the Universal Connectivity Framework is deprecated in Enterprise Data Catalog.
For more information, see the Statement of Support for the Usage of Universal Connectivity Framework (UCF) with Informatica Enterprise Data Catalog.