Zsihovszki Krisztina created NIFI-13487:
-------------------------------------------

             Summary: ConsumeKinesisStream processor does not consume records 
inserted before its start
                 Key: NIFI-13487
                 URL: https://issues.apache.org/jira/browse/NIFI-13487
             Project: Apache NiFi
          Issue Type: Bug
    Affects Versions: 2.0.0-M4
            Reporter: Zsihovszki Krisztina


Noticed a changed behaviour for ConsumeKinesisStream processor when 
"Initial Stream Position" is set to "TRIM_HORIZON".


ConsumeKinesisStream consumes only records which were inserted to Kinesis 
stream after the processor start and it does not consume the records which were 
insterted before processor start.
In case of "TRIM_HORIZON" the records inserted before processor start (app 
registration) should be consumed as well.
 
The processor behaved as it used "LATEST" initial stream position.

The issue occurs for NiFi 2.0 only and the root cause seems to be related to  
NIFI-8531: Upgrading Kinesis processors to AWS SDK 2.x
 
According to 
[https://docs.aws.amazon.com/streams/latest/dev/kcl-migration.html,] setting 
initialPositionInStreamExtended was moved from LeaseManagementConfig to 
RetrievalConfig.

The [existing 
implementation|https://github.com/apache/nifi/blob/main/nifi-extension-bundles/nifi-aws-bundle/nifi-aws-processors/src/main/java/org/apache/nifi/processors/aws/kinesis/stream/ConsumeKinesisStream.java]
 sets the value still in LeaseManagementConfig, this is the reason why the 
default value, "LATEST" was used as initial position.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to