I am currently working with Apache Druid and encountering an issue while attempting to read files from local paths. Specifically, when I try to access files located in a local directory, I receive an error message indicating that there are no files in /opt/druid and the path I have added. I suspect this issue might be related to the need to bind the /opt/druid directory to a corresponding local path on my system. Could you please guide me on how to properly configure this binding?

Issue : 2024-05-16T08:03:08,345 ERROR [qtp475172655-138] org.apache.druid.sql.avatica.DruidMeta – INSERT operations are not supported by requested SQL engine [native], consider using MSQ.
org.apache.druid.error.DruidException: INSERT operations are not supported by requested SQL engine [native], consider using MSQ.

We wanted to use AWS serverless MSK for Druid ingestion but the issue is AWS MSK has hard 200 mbps throughput limit per MSK cluster and our requirement is to have 1800 mbps throughput. Can it be possible that we can add 9 MSK serverless cluster(200*9 = 1800) to ingest data in one Druid datasource. We will ensure upstream that it will ingest unique data to all 9 clusters to avoid any data duplication. I know we can have multiple kafka topics in one MSK cluster to ingest in one Druid datasource using kafka topic pattern, but here the ask is multiple kafka cluster to ingest in one Druid datasource. Is it possible to implement like this in Druid?