[
https://issues.apache.org/jira/browse/KAFKA-6542?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16963776#comment-16963776
]
Matthias J. Sax commented on KAFKA-6542:
----------------------------------------
[~cpreisler], it's unfortunately not that easy. In fact, `StreamThread` will
not process any stream-side records before it goes to state RUNNING. However,
during processing, the table might get updated at any point in time and if
there is out-of-order data, a stream-side record might join with incorrect
table state.
However, the Jira title is miss-leading anyway, because a table update can
never trigger a join computation anyway, because conceptually the stream side
is empty. Hence, a join can only happen if a stream-record is processed.
Also note, if there are no out-of-order data, Kafka Streams ensure that the
table is updated first if the table-record has a smaller timestamp than the
stream-record (cf KAFKA-3514).
Fixing the out-of-order data issue is a larger and more complex change and
thus, would even require a KIP.
> Tables should trigger joins too, not just streams
> -------------------------------------------------
>
> Key: KAFKA-6542
> URL: https://issues.apache.org/jira/browse/KAFKA-6542
> Project: Kafka
> Issue Type: Improvement
> Components: streams
> Affects Versions: 1.1.0
> Reporter: Antony Stubbs
> Priority: Major
>
> At the moment it's quite possible to have a race condition when joining a
> stream with a table, if the stream event arrives first, before the table
> event, in which case the join will fail.
> This is also related to bootstrapping KTables (which is what a GKTable does).
> Related to: KAFKA-4113 Allow KTable bootstrap
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
