You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@spark.apache.org by "Long Sun (JIRA)" <ji...@apache.org> on 2019/04/25 03:21:00 UTC

[jira] [Commented] (SPARK-20462) Spark-Kinesis Direct Connector

    [ https://issues.apache.org/jira/browse/SPARK-20462?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16825669#comment-16825669 ] 

Long Sun commented on SPARK-20462:
----------------------------------

[~gaurav24], [~arushkharbanda],

Any update on this?

"The user does not have to implement their own checkpointing logic and ensuring checkpointing only occurs after the terminal action of a user’s Spark job ensures that no data is lost" The idea above is excellent in the above proposal.

But at least, I doubt why *Spark + Kinesis* don't provide '*offset commit API*' like *Spark+Kafka 0.10.0 or higher.* 

> Spark-Kinesis Direct Connector 
> -------------------------------
>
>                 Key: SPARK-20462
>                 URL: https://issues.apache.org/jira/browse/SPARK-20462
>             Project: Spark
>          Issue Type: New Feature
>          Components: DStreams
>    Affects Versions: 2.1.0
>            Reporter: Lauren Moos
>            Priority: Major
>
> I'd like to propose and the vet the design for a direct connector between Spark and Kinesis. 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org