[
https://issues.apache.org/jira/browse/SPARK-17812?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15563273#comment-15563273
]
Ofir Manor commented on SPARK-17812:
------------------------------------
If you think it is useful, you should definitely propose something that covers
it.
Personally, I can't imagine a case where someone will need that - but you have
much more exposure to different users.
I can only say that it is already covered by the strict "assign" (using only
partitions that were explictly used) - the user will open a new consumer group,
seek to the end, seek specific partitions to its preferred offset and pass the
explictly all the topic-partitions-offsets to the "assign" option
BTW - just from the top of my head - maybe you want'll to add an helper
function or two that:
- gets a specific consumer group, and return whatever object / magic string
that is needed to configure a starting offset from Structured Streaming.
- Same as above, but reads from a Structured Streaming checkpoint and return a
new consumer group (created, seeked, commited). That relates to the "DRP"
ticket, and also just passed to the previous helper function (from checkpoint
to a new streaming job).
> More granular control of starting offsets
> -----------------------------------------
>
> Key: SPARK-17812
> URL: https://issues.apache.org/jira/browse/SPARK-17812
> Project: Spark
> Issue Type: Sub-task
> Components: SQL
> Reporter: Michael Armbrust
>
> Right now you can only run a Streaming Query starting from either the
> earliest or latests offsets available at the moment the query is started.
> Sometimes this is a lot of data. It would be nice to be able to do the
> following:
> - seek back {{X}} offsets in the stream from the moment the query starts
> - seek to user specified offsets
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
