I am going to have about 4 servers (Linux) constantly (constantly as in as many as possible with multiprocessing) send queries (most of the queries with be creating new rows, however one of the servers will read data) to a data server. I imagine it would be better to have separate databases. I think sqlite3 is my best option, but I am open to recommendations. I want to minimize the risk of corruption and data loss any tips?
EDIT: I will also have to check the database(s) to for duplicates.