How can I guarantee data uniqueness in the YTsaurus table?

Question

I have a table with several key columns. The combination of values of those columns should be unique for every row of the table. However, duplicate entries may occur due to a human mistake, and that should be avoided.

What are the best practices here? Are there any built-in options to validate uniqueness or to automatically discard duplicate entries?

Answer 1

In YTsaurus, uniqueness constraints are a part of the schema of a table. Namely, there's the unique_keys attribute of the schema.

Uniqueness constraint is only applied to the table's key. A key is a tuple of key columns. Key columns are those that have their sort_order specified. (The columns constituting the key of a table are always a prefix of the full list of columns.)