[
https://issues.apache.org/jira/browse/STORM-822?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15167902#comment-15167902
]
ASF GitHub Bot commented on STORM-822:
--------------------------------------
Github user hmcl commented on the pull request:
https://github.com/apache/storm/pull/1131#issuecomment-188995178
@revans2 no worries. Concerning your code snippet suggestion, I agree with
most of it. We can definitely keep a lot of the state in the MessageId object.
I agree it is indeed the ideal solution. I considered it initially but was
concerned of how expensive it would be to keep all of that state in the
MessageId.
However, I don't think that Using the markAsDone(id) strategy will suffice
to address the cases where tuples get acked out of order, and the cases were a
sequence of tuples will get acked in sequence, but shifted from the last offset
committed. In this scenario we need to have a way to keep track of the offset
sequences that are ready to be committed. I still think that keeping `acked
Map<TopicPartition, OffsetEntry>` will take care of this in an efficient way.
If there is a better way, of course, I would totally welcome a suggestion.
> As a storm developer I’d like to use the new kafka consumer API (0.8.3) to
> reduce dependencies and use long term supported kafka apis
> --------------------------------------------------------------------------------------------------------------------------------------
>
> Key: STORM-822
> URL: https://issues.apache.org/jira/browse/STORM-822
> Project: Apache Storm
> Issue Type: Story
> Components: storm-kafka
> Reporter: Thomas Becker
> Assignee: Hugo Louro
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
