You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tubemq.apache.org by "Guocheng Zhang (Jira)" <ji...@apache.org> on 2020/04/11 05:13:00 UTC
[jira] [Comment Edited] (TUBEMQ-62) consumed and set
consumerConfig.setConsumeModel (0) for the first time, found that the
consumed data contains the previous data
[ https://issues.apache.org/jira/browse/TUBEMQ-62?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17081140#comment-17081140 ]
Guocheng Zhang edited comment on TUBEMQ-62 at 4/11/20, 5:12 AM:
----------------------------------------------------------------
setConsumeModel() is used to define the offset behavior when the consumer group starts. The analysis found that there is a scenario that is not covered during the first registration: if some consumers exit at startup, the allocated partitions are distributed to other consumers, because these partitions are not for the first time, other consumers will think that the corresponding partition has been set up to use the continuous consumption method for consumption. In fact, these partitions have not yet started to consume, so the data is more than expected.
like this:
!image-2020-04-11-13-11-31-216.png!
!image-2020-04-11-13-12-13-520.png!
was (Author: gosonzhang):
setConsumeModel() is used to define the offset behavior when the consumer group starts. The analysis found that there is a scenario that is not covered during the first registration: if some consumers exit at startup, the allocated partitions are distributed to other consumers, because these partitions are not for the first time, other consumers will think that the corresponding partition has been set up to use the continuous consumption method for consumption. In fact, these partitions have not yet started to consume, so the data is more than expected.
like this:
!first.png!
!second.png!
> consumed and set consumerConfig.setConsumeModel (0) for the first time, found that the consumed data contains the previous data
> -------------------------------------------------------------------------------------------------------------------------------
>
> Key: TUBEMQ-62
> URL: https://issues.apache.org/jira/browse/TUBEMQ-62
> Project: Apache TubeMQ
> Issue Type: Bug
> Reporter: tiyeliyameng
> Assignee: Guocheng Zhang
> Priority: Major
> Attachments: image-2020-04-11-13-11-31-216.png, image-2020-04-11-13-12-13-520.png, second.png
>
>
> I applied for a new tube and set consumerConfig.setConsumeModel (0). Then I produced 30000 rows , but the consumption reached 140000, and some old data had been consumed.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)