You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kafka.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2016/05/05 04:53:12 UTC
[jira] [Commented] (KAFKA-3659) Consumer does not handle
coordinator connection blackout period gracefully
[ https://issues.apache.org/jira/browse/KAFKA-3659?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15271886#comment-15271886 ]
ASF GitHub Bot commented on KAFKA-3659:
---------------------------------------
GitHub user hachikuji opened a pull request:
https://github.com/apache/kafka/pull/1322
KAFKA-3659: Handle coordinator disconnects more gracefully in client
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/hachikuji/kafka KAFKA-3659
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/kafka/pull/1322.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #1322
----
commit e3d0f3f950fc7005e690510b0f66a19d4ca4c10c
Author: Jason Gustafson <ja...@confluent.io>
Date: 2016-05-05T00:33:36Z
KAFKA-3659: Handle coordinator disconnects more gracefully in client
----
> Consumer does not handle coordinator connection blackout period gracefully
> --------------------------------------------------------------------------
>
> Key: KAFKA-3659
> URL: https://issues.apache.org/jira/browse/KAFKA-3659
> Project: Kafka
> Issue Type: Bug
> Components: consumer
> Affects Versions: 0.9.0.0, 0.9.0.1
> Reporter: Jason Gustafson
> Assignee: Jason Gustafson
>
> Currently when the connection to the coordinator is closed, the consumer will immediately try to rediscover the coordinator and reconnect to it. This is fine as it is, but the NetworkClient enforces a blackout period before it will allow the reconnect to be attempted. This causes the following cycle which continues in a fairly tight loop until the blackout period has completed:
> 1. Notice connection failure (i.e. DISCONNECTED state in ConnectionStates)
> 2. Send GroupCoordinator request to rediscover coordinator.
> 3. Attempt to connect to coordinator.
> 4. Go back to 1.
> To fix this, we should avoid rediscovery while the connection is blacked out.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)