You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flink.apache.org by "CHEN Zhongpu (Jira)" <ji...@apache.org> on 2023/02/26 14:05:00 UTC

[jira] [Created] (FLINK-31228) flink-playgrounds is using a buggy zookeeper

CHEN Zhongpu created FLINK-31228:
------------------------------------

             Summary: flink-playgrounds is using a buggy zookeeper
                 Key: FLINK-31228
                 URL: https://issues.apache.org/jira/browse/FLINK-31228
             Project: Flink
          Issue Type: Bug
            Reporter: CHEN Zhongpu


The flink-playgrounds [1] is using a buggy zookeeper image from docker hub, and thus the kafka can never connect to the zookeeper server.
----
How to reproduce the bug?

For example, after running "docker-compose up -d" in operations-playground [2], we can observe in Flink web UI that the job will in "restarting" state soon, and both "zookeeper" and "kafka" are stopped.
----
I noticed that the used "wurstmeister/zookeeper:3.4.6" image was published in 7 years ago, and it is badly documented. In my test, switching to "bitnami/zookeeper:3.7.1" can solve this problem.

Also, I think it would be better to discard images published by wurstmeister, since they are not updated frequently (and there buggy), and we should switch to bitnami.

 

 [1][https://github.com/apache/flink-playgrounds|https://github.com/apache/flink-playgrounds)]

 [2] [https://github.com/apache/flink-playgrounds/tree/master/operations-playground)]



--
This message was sent by Atlassian Jira
(v8.20.10#820010)