Which approach meets the requirement for a centralized metadata layer?

A company has several teams of analytics. Each team of analysts has their own cluster. The teams need to run SQL queries using Hive, Spark-SQL and Presto with Amazon EMR. The company needs to enable a centralized metadata layer to expose the Amazon S3 objects as tables to the analysts.

Which approach meets the requirement for a centralized metadata layer?
A . EMRFS consistent view with a common Amazon DynamoDB table
B . Bootstrap action to change the Hive Metastore to an Amazon RDS database
C . s3distcp with the output Manifest option to generate RDS DDL
D . naming scheme support with automatic partition discovery from Amazon S3

Answer: B

Explanation:

https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-metastore-external-hive.html