You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@kafka.apache.org by "Jun Rao (JIRA)" <ji...@apache.org> on 2016/02/02 17:05:39 UTC

[jira] [Commented] (KAFKA-2936) Socket server selector can stuck on one send in tight loop.

    [ https://issues.apache.org/jira/browse/KAFKA-2936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15128448#comment-15128448 ] 

Jun Rao commented on KAFKA-2936:
--------------------------------

[~becket_qin], is this still an issue? We fixed a similar issue in KAFKA-2813.

> Socket server selector can stuck on one send in tight loop.
> -----------------------------------------------------------
>
>                 Key: KAFKA-2936
>                 URL: https://issues.apache.org/jira/browse/KAFKA-2936
>             Project: Kafka
>          Issue Type: Bug
>          Components: network
>    Affects Versions: 0.9.0.0
>            Reporter: Jiangjie Qin
>            Assignee: Onur Karaman
>             Fix For: 0.9.0.1
>
>
> When broker was sending a FetchResponse it is possible that the data to be sent back is truncated. In this case, a KafkaException will be thrown. This exception is caught by processor and the selector will be sending the message in a tight while loop. It will continue doing this until the socket is closed by the client due to request timeout.
> We should have a max retry for the response send.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)