[jira] [Commented] (KAFKA-3184) Add Checkpoint for In-memory State Store

["Matthias J. Sax (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KAFKA-3184?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16701422#comment-16701422 ] 

Matthias J. Sax commented on KAFKA-3184:
----------------------------------------

Just a side comment: RocksDB allows to "pin/freeze" a current state to protect it from compaction. This is a light weight metadata operation. After a version is pinned (not sure atm what the correct RocksDB term for this operation is), the immutable SST files can be copied in the background to complete a checkpoint.

> Add Checkpoint for In-memory State Store
> ----------------------------------------
>
>                 Key: KAFKA-3184
>                 URL: https://issues.apache.org/jira/browse/KAFKA-3184
>             Project: Kafka
>          Issue Type: Improvement
>          Components: streams
>            Reporter: Guozhang Wang
>            Priority: Major
>              Labels: user-experience
>
> Currently Kafka Streams does not make a checkpoint of the persistent state store upon committing, which would be expensive since it is "stopping the world" and write on disks: for example, RocksDB would require you to copy the file directory to make a copy naively. 
> However, for in-memory stores checkpointing maybe doable in an asynchronous manner hence it can be done quickly. And the benefit of having intermediate checkpoint is to avoid restoring from scratch if standby tasks are not present.



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