[jira] [Updated] (STORM-2432) Storm-Kafka-Client Trident Spout Seeks Incorrect Offset With UNCOMMITTED_LATEST Strategy

["Raghav Kumar Gautam (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/STORM-2432?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Raghav Kumar Gautam updated STORM-2432:
---------------------------------------
    Attachment: worker_works.log.zip
                worker_problem.log.zip

> Storm-Kafka-Client Trident Spout Seeks Incorrect Offset With UNCOMMITTED_LATEST Strategy
> ----------------------------------------------------------------------------------------
>
>                 Key: STORM-2432
>                 URL: https://issues.apache.org/jira/browse/STORM-2432
>             Project: Apache Storm
>          Issue Type: Bug
>          Components: storm-kafka-client
>    Affects Versions: 1.x
>            Reporter: Raghav Kumar Gautam
>            Assignee: Hugo Louro
>             Fix For: 1.x
>
>         Attachments: worker_problem.log.zip, worker_works.log.zip
>
>
> With UNCOMMITED_LATEST offset, Storm-Kafka-Client Trident Spout doesn't read all the data.
> See worker logs for of the same topology with UNCOMMITED_EARLIEST as well as UNCOMITTED_LATEST.
> The source topic has 3 partition and I publish data k000000-k000499(k000000, v000000 etc.)
> In particular for k000000 value the data is not picked by spout at all.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)