[
https://issues.apache.org/jira/browse/FLINK-31108?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17718508#comment-17718508
]
Danny Cranmer commented on FLINK-31108:
---------------------------------------
Merged commit
[{{a62e3d8}}|https://github.com/apache/flink-connector-aws/commit/a62e3d866338f71345a92ac9d7e056eba603a763]
into apache:main
> Use StreamARN for API calls in Kinesis Connector
> ------------------------------------------------
>
> Key: FLINK-31108
> URL: https://issues.apache.org/jira/browse/FLINK-31108
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / Kinesis
> Affects Versions: 1.15.3, aws-connector-4.1.0
> Reporter: Hong Liang Teoh
> Assignee: Hong Liang Teoh
> Priority: Major
> Labels: pull-request-available
> Fix For: aws-connector-4.2.0
>
>
> Currently, the FlinkKinesisConsumer (Polling + EFO) + FlinkKinesisProducer
> uses the stream name during API calls
> We want to change this to the StreamARN. There are two reasons for this:
> - This allows lower latency calls to the Kinesis endpoint for GetRecords API
> - Paves the way for allowing user target cross-account streams without
> assume role (i.e. IAM role in account A but target stream in account B)
>
> The APIs that are currently called:
> * ListShards
> * GetShardIterator
> * GetRecords
> * DescribeStream
> * DescribeStreamSummary
> * DescribeStreamConsumer (already uses StreamARN)
> * RegisterStreamConsumer (already uses StreamARN)
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
