Hi Flink devs,
I am currently working on a AWS Kinesis streaming connector for Flink, and have filed a JIRA ticket for the feature:
https://issues.apache.org/jira/browse/FLINK-3211.
As described in the JIRA description, after a close look at the AWS SDK and AWS KCL (Kinesis Client Library) / KPL (Kinesis Producer Library), I have currently chosen to go for AWS SDK for the source and sink implementation.
I particularly need some feedback on the parts where the Kinesis connector will differ from the Kafka connector implementation, listed in the JIRA.
Any thoughts will be helpful, thanks!