Release 19.1.0.0.6_MLR33251626

Bug 33172932: MISSING TUPLES WHEN PIPELINE DEPLOYED ON MULTIPLE EXECUTORS

Fixed the missing tuples issue, when a pipeline is deployed on multiple executors.

Bug 33193157: DRUID CUBE CREATION FAILS WHEN KAFKA TARGET USES A KAFKA CONNECTION CONFIGURED WITH BOOTSTRAP SERVER

Fixed the Druid cube creation failure issue, when a Kafka target uses a Bootstrap Kafka connection.

Bug 32964239: DIFFERENTIATE PIPELINE TERMINATION BETWEEN SOFT EXCEPTION AND RUNTIME EXCEPTION

Issue

The pipeline terminates due to incompatible input data

Fix

Based on your requirement, you can terminate the pipeline, or continue by dropping the erroneous data. Update the osa.continueon.data.exception flag in the OSA_SYSTEM_PROPERTY table:

If the flag is set to true, the pipeline will continue to run, dropping the faulty event.
If the flag is set to false, pipeline will terminate.
By default the value is set to true.

Bug 32964187: CONFIGURATION PARAMETER ISSUE TO ALLOW PIPELINE TO CONTINUE OR TERMINATE ON EXCEPTION

Issue

The pipeline terminates due to incompatible input data

Fix

Based on your requirement, you can terminate the pipeline, or continue by dropping the erroneous data. Update the osa.continueon.data.exception flag in the OSA_SYSTEM_PROPERTY table:

If the flag is set to true, the pipeline will continue to run, dropping the faulty event.
If the flag is set to false, pipeline will terminate.
By default the value is set to true.

Bug 32990445: IGNITE SERVER IS CALCULATING WRONG NUMBER OF WORKER NODES.

Fixed the ignite server calculating wrong number of worker nodes issue.

Bug 33208855: CANNOT REPUBLISH A PIPELINE WITH OFFSET THAT WAS KILLED DUE TO INPUT DATA EXCEPTION

Fixed issue related to republishing a pipeline that has an erroneous data at the beginning of the offset.

Bug 33236693: ISSUE WITH SYSTEM PROPERTY "SPARK.STREAMING.EXCEPTION.ISTERMINATE" SET TO FALSE

Fixed the issue in which setting spark.streaming.exception.isTerminate in the OSA_SYSTEM_PROPERTY table to FALSE does not keep the pipeline running when exceptions are encountered.

Bug 33272967: ENTIRE BATCH IS DROPPED WHEN THERE IS A FAULTY EVENT IN THE BATCH

Issue

When spark.streaming.exception.isTerminate is set to false, and if any record failed to get inserted into the Oracle DB table, then the entire batch is skipped.

Fix

You can skip only the erroneous data that is failing insertion, by updating the osa.jdbc.switchtononbatch.insert to true, in the OSA_SYSTEM_PROPERTY table.