(Ported from memsql-public-chat Slack channel)
Bhargava [7:33 AM]
Hi guys, I’m trying to understand the indexes for rowstore and columnstore tables, rowstore table restricts duplicate rows with the same primary keys but columnstore has duplicate rows with duplicate primary keys, how can we restrict columnstore tables to only accept rows with unique keys ?
alec [7:38 AM]
@Bhargava columnstore has unenforced unique keys (SingleStoreDB Cloud · SingleStore Documentation) if you can enforce the constraint via your application, or you can use rowstore and insert/select to columnstore in a stored proc, for example. for columnstore architecture and use cases see SingleStoreDB Cloud · SingleStore Documentation, the main concept is sorting your data via clustered columnstore index (in addition to shard key) (edited)
Bhargava [7:39 AM]
Thanks Alec !