kafka-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jay Kreps (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (KAFKA-45) Broker startup, leader election, becoming a leader/follower for intra-cluster replication
Date Fri, 09 Mar 2012 21:57:00 GMT

    [ https://issues.apache.org/jira/browse/KAFKA-45?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13226523#comment-13226523
] 

Jay Kreps commented on KAFKA-45:
--------------------------------

I think we are overthinking this. Currently cleanup is not a precise SLA, it is just a guarantee
of the form "we will never delete anything younger than X OR we will always maintain at least
Y bytes of messages". Trying to maintain this in synchronous form across nodes is overkill
I think. It is fine if every node acts independently as long as each of them respects the
SLA. I think this should be much simpler and more likely to work.
                
> Broker startup, leader election, becoming a leader/follower for intra-cluster replication
> -----------------------------------------------------------------------------------------
>
>                 Key: KAFKA-45
>                 URL: https://issues.apache.org/jira/browse/KAFKA-45
>             Project: Kafka
>          Issue Type: Bug
>            Reporter: Jun Rao
>
> We need to implement the logic for starting a broker with replicated partitions, the
leader election logic and how to become a leader and a follower.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message