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

Posted to dev@kafka.apache.org by "Pradeep Bansal (JIRA)" <ji...@apache.org> on 2019/02/01 05:57:00 UTC

[jira] [Created] (KAFKA-7891) Reduce time to start kafka server with clean state

Pradeep Bansal created KAFKA-7891:
-------------------------------------

             Summary: Reduce time to start kafka server  with clean state
                 Key: KAFKA-7891
                 URL: https://issues.apache.org/jira/browse/KAFKA-7891
             Project: Kafka
          Issue Type: Improvement
    Affects Versions: 2.1.0
            Reporter: Pradeep Bansal


I am using kafka 2.1.0 and has 6 broker cluster. In this I had a scenario where a topic (with replication factor 3 and min insync replica as 2) leader went down and we lost its data. When starting this broker with fresh state, it took aroun 45 minutes to catch up with replicas (data size of 190 G).

 

Is it expected to take such a large period to recover. are there configuration using which this can be optimized.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)