You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Jose Armando Garcia Sancio (Jira)" <ji...@apache.org> on 2021/03/09 02:20:00 UTC
[jira] [Comment Edited] (KAFKA-10817) Add clusterId validation to
Fetch handling
[ https://issues.apache.org/jira/browse/KAFKA-10817?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17297787#comment-17297787 ]
Jose Armando Garcia Sancio edited comment on KAFKA-10817 at 3/9/21, 2:19 AM:
-----------------------------------------------------------------------------
{quote}ClusterId validation is only applied to Fetch requests? should we also apply it to BeginQuorum EndQuorum FetchSnapshot?
{quote}
Yes. [KIP-595|https://cwiki.apache.org/confluence/display/KAFKA/KIP-595%3A+A+Raft+Protocol+for+the+Metadata+Quorum] describes that Vote, BeginQuorum and EndQourum specify and validate the clusterId. I created https://issues.apache.org/jira/browse/KAFKA-12440
was (Author: jagsancio):
> ClusterId validation is only applied to Fetch requests? should we also apply it to BeginQuorum EndQuorum FetchSnapshot?
Yes. [KIP-595|https://cwiki.apache.org/confluence/display/KAFKA/KIP-595%3A+A+Raft+Protocol+for+the+Metadata+Quorum] describes that Vote, BeginQuorum and EndQourum specify and validate the clusterId. I created https://issues.apache.org/jira/browse/KAFKA-12440
> Add clusterId validation to Fetch handling
> ------------------------------------------
>
> Key: KAFKA-10817
> URL: https://issues.apache.org/jira/browse/KAFKA-10817
> Project: Kafka
> Issue Type: Sub-task
> Reporter: Jason Gustafson
> Assignee: David Jacot
> Priority: Major
>
> Initially we were unsure how the clusterId would be generated by the cluster after Zookeeper removal, so we did not implement it. It is looking now like we will probably require users to generate it manually prior to starting the cluster. See here for details: https://cwiki.apache.org/confluence/display/KAFKA/KIP-631%3A+The+Quorum-based+Kafka+Controller#KIP631:TheQuorumbasedKafkaController-Fencing. In this case, we can assume that the clusterId will be provided when instantiating the raft client, so we can add the logic to the request handler to validate it in inbound Fetch requests.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)