Stanislav Knot

Is it possible to provide logs at DEBUG level? If I am reading the logs correctly, you created the `oceana.interval.topic` topic (it seems to be created in the Kafka too),...

I just tried with your configuration and the users were ready in range of seconds. I will do more testing to see if I can get the same error as...

@GarimaBathla I did further testing with 500 KafkaUsers. I did find out that each KafkaUser does require 5 AdminClient calls each reconciliation. That means high number of KUs is basically...

I used the same configuration as you posted above. I think it is very flaky.

I tried batching for quotas and I was able to get 340 ready KafkaUsers. For comparison, without batching I can get ~80 ready KafkaUsers.

I created a PoC. The merging of reconciliation result for TLS and SCRAM user needs to be improved https://github.com/sknot-rh/strimzi/commit/8719e87d8987801295a0b091ae60feb28ed7aec1#diff-9c42d639637a67deb0200524f18b23daf141adeddc47f4d8d04145c76ed97995R61

Yeah, creating the users with some delays does not help IMO. The issue is, that AdminClient sends _numberOfUsers*5_ requests (200 users means 1000 requests) to brokers each reconciliation. So after...

@scholzj Is this a valid issue for us? AFAIU we are moving away from SS.

I think we are doing some basic validation for Kafka resource (https://github.com/strimzi/strimzi-kafka-operator/blob/main/cluster-operator/src/main/java/io/strimzi/operator/cluster/model/KafkaCluster.java#L384). IIRC there is no validation in the KafkaTopic `spec.config`.

I think `io.strimzi.operator.common.AbstractOperator#validate` does validate the entire `spec` but not the `config` (since it is `Map`). However the `spec.config` validation is done in `io.strimzi.operator.cluster.model.KafkaConfiguration#validate`. It reads config model and for...