Understanding Kafka Indexing Service parallelism

Hi Team,

I am running kafka Indexing service and seems it is working fine. However, I noticed that only one task is showing the offset correctly, rest other tasks are showing both starting and current offset as same (the value is 2 ) . My assumption was all partitions will be consumed in parallel as I see 16 Active tasks running (8 topic partitions * 2 replica). Would anyone help me understand if I am missing any here. Here is my ioConfig from Supervisor spec.

},“ioConfig”: {

1. ```
       "topic": "kafka.druid.topic",
  1.       "consumerProperties": {
    
1. ```
       "bootstrap.servers": "kafkabroker1.com:6667,kafkabroker2:6667,kafkabroker3:6667,kafkabroker4.com:6667",
  1.       "security.protocol": "SASL_PLAINTEXT",
    
1. ```
       "sasl.kerberos.service.name": "kafka",
  1.       "group.id": "kafka-indexer2",
    
1. ```
       "linger.ms": "2"
  1.   },
    
1. ```
   "taskCount": 8,
  1.   "replicas": 2,
    
1. ```
   "taskDuration": "PT1H"

And here is what I see in overlord console.

Hi Kashif,

Druid does attempt to consume the partitions in parallel. Are you sure you are producing data to all of the partitions?