I have a very simple database, which consists of two tables:
TabA , idA binary(32) Primarykey NOT NULL
TabB, idB binary(32) Primarykey NOT NULL
Only two simple activities are performing on these tables and they are independent.
Inserting new id to idA in TabA.
Searching TabB idB column for a given value, if not exists insert the given value.
The two tables could grow into billions of records. The data can’t be housekept. Currently the database is in a relational database MS SQL.
Based on the very simple query and flat schema, which type of databases could provide best performance? Search engine? In memory KV? Mapreduce? Thanks in advance!