[ https://issues.apache.org/jira/browse/SPARK-21233?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16066088#comment-16066088 ]
Sean Owen commented on SPARK-21233: ----------------------------------- Where would you put it instead? Kafka already provides for storing offsets in Kafka. This sounds like a big change and I don't see detail or design here. > Support pluggable offset storage > -------------------------------- > > Key: SPARK-21233 > URL: https://issues.apache.org/jira/browse/SPARK-21233 > Project: Spark > Issue Type: New Feature > Components: DStreams > Affects Versions: 2.0.2, 2.1.1 > Reporter: darion yaphet > > Currently we using *ZooKeeper* to save the *Kafka Commit Offset* , when there > are a lot of streaming program running in the cluster the ZooKeeper Cluster's > loading is very high . Maybe Zookeeper is not very suitable to save offset > periodicity. > This issue is wish to support a pluggable offset storage to avoid save it in > the zookeeper -- This message was sent by Atlassian JIRA (v6.4.14#64029) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org