[
https://issues.apache.org/jira/browse/KAFKA-2390?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14709607#comment-14709607
]
Jay Kreps commented on KAFKA-2390:
----------------------------------
Thanks for the clear problem statement, I agree with the problem. I thought we
already discussed the fix for this and agreed on just adding the partition(s)
and offset(s) that are out of range to the existing OffsetOutOfRangeException?
The key point is that it is actually quite ambiguous whether the seek "caused"
the out of range fetch or not. The seek may actually be valid at seek time, but
become invalid later when the fetch occurs. I agree that you need to know you
are out of range but figuring out whether this was "caused" by the seek or not
is actually not necessarily doable and I don't think actually helps.
I propose the fix as just taking the fetch offset we already have per-partition
and including this with the partition in the exception that is thrown. Thoughts?
> Seek() should take a callback.
> ------------------------------
>
> Key: KAFKA-2390
> URL: https://issues.apache.org/jira/browse/KAFKA-2390
> Project: Kafka
> Issue Type: Sub-task
> Reporter: Jiangjie Qin
> Assignee: Dong Lin
>
> Currently seek is an async call. To have the same interface as other calls
> like commit(), seek() should take a callback. This callback will be invoked
> if the position to seek triggers OFFSET_OUT_OF_RANGE exception from broker.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
