Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

Kindly be informed that starting with SNP Glue version 2408, the SNP Java Connector now requires JRE 11.

In prior releases, the minimum required version was JRE 8; however, this has changed, and JRE 11 is now the minimal required version.

Please note that SNP plans to decommission the Enable Merge feature in the SNP Glue extraction process. Starting with release 2502, this feature will no longer be configurable and will be fully removed in release 2511. 

The Enable Merge feature was originally introduced to perform delta updates on connected data storage. However, its direct integration with the load process has proven inefficient. We now offer more efficient methods for performing merges by decoupling the load and merge processes, utilizing platform-based solutions such as our /wiki/spaces/GLUE/pages/4288151553. This approach not only improves performance but also reduces credit consumption.

As a result, we have decided to transition to these improved functionalities and decommission the Enable Merge feature. This change affects only Enable Merge and does not impact the Enable Update feature or customers replicating to SQL databases.

Improvements

Extractors

  • Continuing with the enhancements of Parallel processing (mass execution), Glue table fetchers, HANA fetchers, and LISTCUBE fetchers allow the use of all available split options for parallel replication.

  • The procedure for replicating the STXL table has been improved. This allows parallel execution that optimizes the whole process and decreases the execution time.

  • In the Request Monitor, in rare instances, certain Glue requests may remain in a running or pending status even after they have been processed. This may negatively affect the performance of the monitor. In the latest release, we have introduced a report that consolidates statuses and thus solves potential performance problems.

CDC

  • The latest release introduces a new setting, enabling you to define the trigger name pattern. All triggers created by SNP Glue will adhere to this specified pattern. For further details, please refer to our documentation.

  • In SNP Glue release 2405, a new feature was introduced allowing records to be marked as archived (A) rather than deleted. Initially implemented for the HANA database, this functionality is now also available for the Oracle database. In the latest release, the functionality has been enabled for cluster tables on the HANA database.

  • SNP Glue trigger management now includes a trigger repair functionality, designed to resolve any inconsistencies that may arise with triggers.

SLT

  • The SNP Glue SLT Integration Add-On now includes a setting that allows placeholders to control the target table name. This feature enables the definition of container names for file storage or target table names for database storage.

  • Additionally, SNP Glue now supports data replication from SAP tables with long table names. Previously, this capability was restricted to table names of up to 26 characters. The latest release has enhanced the replication logic to support tables with extended name lengths.

Storage Management

  • The SNP Java Connector versioning has been synchronized with SNP Glue versioning. As a result, SNP Glue release 2411 now requires SNP Java Connector version 2411.

  • Google BigQuery tables created by SNP Glue now include primary key information.

  • Google Cloud Cortex integration has been enabled for the Google BigQuery Streaming storage.

  • Several enhancements have been implemented for AWS Redshift storage. The parameters JDBC call repeat and JDBC repeat delay have been introduced to enable automatic retries of JDBC calls in the event of operation failures. Users can now define the distribution style for Redshift tables; previously, this was automatically set to AUTO. Furthermore, it is now possible to specify the STS endpoint when utilizing the IAM authentication method.

Fixes

Data Dictionary

  • Previously, fields of the RAW type in source tables were automatically mapped to the String type. This behavior has been adjusted so RAW fields are now mapped to the dedicated data element /DVD/GL_RAW_STRG. The data type in the target table can now be controlled through data element mapping for the specified storage, allowing for mapping to the BINARY data type when necessary.

  • Reactivation of a Glue table was previously unsupported when binary storage was used, and the specified container name differed from the Glue table name.

Extractors

  • The key deduplication process for SAP cluster tables was not functioning correctly, which could have led to duplicate data. This logic has now been corrected and is functioning as expected.
    Also, the deduplication logic for DSO records has been corrected.

  • Previously, during replication to file storage, no CSV file was generated in cases of an empty load. This functionality has now been corrected, and a CSV file is created even when it contains no data.

  • During periods of heavy load, some jobs had the potential to hang due to a locking mechanism within a background job responsible for collecting and compressing load statistics. The root cause was identified, and the locking mechanism has been modified to prevent interference with active extraction jobs, ensuring smoother performance during peak usage.

  • Replication to AVRO files was failing because of an incorrect mapping of currency fields

TMS

  • The Glue Transport Management System previously allowed only the transfer of active objects between systems, preventing the transfer of Glue objects in a saved status. This limitation has now been resolved, enabling the transfer of objects in saved status as well.

Storage management

  • In rare circumstances, replication to Snowflake could lead to data inconsistencies. This issue required multiple network failures occurring in a specific sequence. The logic has now been improved to prevent these scenarios, ensuring data consistency in all cases.

  • The storage management configuration did not function correctly if a user lacked permission to read RFC in SM59. The logic has now been updated to display an error message indicating that permission adjustments are required for proper functionality.

  • No labels