I am fairly new to hadoop ecosystem, so any advice is appreciated. I am working on creating a customer360 database for BI purpose on my mapr environment. This requires inserts and updates and may be even deletes (rarely).
Is it a good idea to turn on ACID in Hive and create the entire schema in Hive ? Or is it a good idea to create the table that needs updates and deletes in Hbase and then have a hive on hbase concept . I plan to use the hive and the hive on hbase for ETL and then use Drill to read the tables (hive and hbase) .
I general is it a good idea to turn on transaction (ACID) in Hive ?
Thanks for you help.