[ 
https://issues.apache.org/jira/browse/KAFKA-8827?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

frid updated KAFKA-8827:
------------------------
    Description: 
i have kafka cluster with version 0.10.2, 3 brokers an 3 zookeeper i try to 
generete 25 mb/sec to my topic wich have 400 partitions, the cluster behavior 
is very strange

the followers stop receiving messages each 5 min but the leader still receive 
messages, which generates lag so underreplicated partitions and the cluster 
become unstable. cpu reache 80% in 2 brokers but in the first still in 20 %

Could someone explain this phenomenon?  thanks in advance find here some metric

!image-2019-08-23-15-35-17-558.png!

  was:
i have kafka cluster with version 0.10.2, 3brokers an 3 zookeeper i try to 
generete 25 mb/sec to my topic wich have 400 partitions, the cluster behavior 
is very strange

the followers stop receiving messages each 5 min but the leader still receive 
messages, which generates lag so underreplicated partitions and the cluster 
become unstable. cpu reache 80% in followers brokers but in leadear still in 20 
%

Could someone explain this phenomenon?  thanks in advance find here some metric

!image-2019-08-23-15-35-17-558.png!

        Summary: kafka brokers intermittent traffic  (was: kafka followers 
intermittent traffic)

> kafka brokers intermittent traffic
> ----------------------------------
>
>                 Key: KAFKA-8827
>                 URL: https://issues.apache.org/jira/browse/KAFKA-8827
>             Project: Kafka
>          Issue Type: Bug
>    Affects Versions: 0.10.2.0
>            Reporter: frid
>            Priority: Critical
>         Attachments: image-2019-08-23-15-35-17-558.png
>
>
> i have kafka cluster with version 0.10.2, 3 brokers an 3 zookeeper i try to 
> generete 25 mb/sec to my topic wich have 400 partitions, the cluster behavior 
> is very strange
> the followers stop receiving messages each 5 min but the leader still receive 
> messages, which generates lag so underreplicated partitions and the cluster 
> become unstable. cpu reache 80% in 2 brokers but in the first still in 20 %
> Could someone explain this phenomenon?  thanks in advance find here some 
> metric
> !image-2019-08-23-15-35-17-558.png!



--
This message was sent by Atlassian Jira
(v8.3.2#803003)

Reply via email to