Multi tenancy using shared data sources by kafka indexing service

Hi Druids

We are using druid 0.15 and we want to achieve multi tenancy using shared data sources technique for a single Kafka topic.

Kafka indexing service is being used to ingest data into druid. A string value for the dimension tenant_id which is unique for an application is ingested.

The following tuning configuration is specified in the ingestion spec.

  "maxRowsPerSegment": 5000000,
  "maxRowsInMemory": 25000,
  "patitionsSpec": {
    "type": "kafka",
    "numShards": 5,
    "partitionDimensions": ["tenant_id"],
    "assumeGrouped": true

We are not able to achieve multi tenancy as same row is getting ingested into multiple datasources which are using single Kafka topic. Can we achieve multi tenancy using Kafka indexing service?