You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Mahesh Madushan (Jira)" <ji...@apache.org> on 2023/05/11 08:59:00 UTC
[jira] [Commented] (KAFKA-13349) Allow Iterator.remove on KeyValueIterator
[ https://issues.apache.org/jira/browse/KAFKA-13349?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17721696#comment-17721696 ]
Mahesh Madushan commented on KAFKA-13349:
-----------------------------------------
Hello [~guozhang] . While I'm looking into this issue , and I'm new to the codebase and wondering that the range iterator mention here is a KeyValueIterator and if we want to support remove, we should implement remove method in all implementation of the KeyValueIterator? is my thought process is correct ? Thank You.
> Allow Iterator.remove on KeyValueIterator
> -----------------------------------------
>
> Key: KAFKA-13349
> URL: https://issues.apache.org/jira/browse/KAFKA-13349
> Project: Kafka
> Issue Type: Improvement
> Components: streams
> Reporter: Guozhang Wang
> Priority: Major
> Labels: needs-kip, newbie++
>
> Today Stream's state store's range iterator does not support `remove`. We could consider adding such support for all the built-in state stores:
> * RocksDB's native iterator does not support removal, but we can always do a delete(key) concurrently while the iterator is open on the snapshot.
> * In-Memory: straight forward implementation.
> The benefit of that is then for range-and-delete truncation operation we do not necessarily have to be cautious about concurrent modification exceptions. This could also help GC with in-memory stores.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)