jsancio commented on a change in pull request #10289:
URL: https://github.com/apache/kafka/pull/10289#discussion_r595378358
##########
File path: raft/src/main/java/org/apache/kafka/raft/KafkaRaftClient.java
##########
@@ -939,12 +951,27 @@ private FetchResponseData buildEmptyFetchResponse(
);
}
- private boolean hasValidClusterId(FetchRequestData request) {
+ private boolean hasValidClusterId(ApiMessage request) {
Review comment:
Yes. @dengziming in that example, the user has incorrectly configured
the cluster. The user was configured it so that all of the controllers have
each other's listener (connection) information but the cluster ids are
different.
The question is do we want to catch those misconfiguration early by shutting
down the brokers/controllers? Or do we want to continue executing with the user
potentially missing that the controllers/brokers are incorrectly configuration?
There have been conversation of having the first controller leader generate
the cluster id and replicate that information to all off the nodes. The
currently implementation generate the cluster id in the `StorateTool` which the
user has to run when configuring the controllers.
I am okay leaving it as is and addressing this in a future PR.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
