I’m having trouble with this situation. First I tried to migrate the volume which is in zookeeper to new kraft cluster with MirrorMaker2 but I couldn’t manage it. I posted that as a question as well. Now I tried to follow the migration steps that is described in Confluent’s Docs. But I couldn’t manage this also.

We recently started migrating our test clusters from Zookeeper to KRaft, all running Confluent Kafka on prem. In KRaft, cluster’s metadata is stored in __cluster_metadata topic (see: https://developer.confluent.io/courses/architecture/control-plane/#kraft-cluster-metadata) . What I noticed after the first cluster was migrated is that I am not able to describe this topic using kafka-topics command, it just does not exist:

I have a kafka producer configured with at least once delivery semantics. The challenge now is How do I actually Implement a scenario to force Kafka to retry and publish Duplicate messages? The following are some of the ideas that came to mind —- Any better ideas?

We have java sink connectors recently upgraded kafka connect from 5.3.1 to 7.3.1.

org.apache.kafka.common.config.ConfigException: Invalid value [] for configuration bootstrap.servers: Empty list
at io.confluent.connect.utils.validators.NonEmptyListValidator.ensureValid(NonEmptyListValidator.java:21)

While starting the server for confluent kafka in windows 10, getting this error
Error: Could not find or load main class org.apache.zookeeper.server.quorum.QuorumPeerMain confluent kafka

Im building a streaming application where I will receive logs of various transactions which need to be processed. Planning to use kafka for this, and Im a newbie to kafka. What are the best practices when choosing a topic for kafka, are they supposed to be static values? Also in my case, each log will have a transaction-id. Can I use the value of transaction-id to generate a new topic when a new unique value of transaction-id is published to kafka?

I am learning the designs of Kafka replication, and producers.
And to my surprise, I somehow get the conclusion that it is impossible for producers to know accurately if the messages are committed by the broker, and the safety of writes are agnostic.

I am learning the designs of Kafka replication, and producers.
And to my surprise, I somehow get the conclusion that it is impossible for producers to know accurately if the messages are committed by the broker, and the safety of writes are agnostic.