Ingest pre-horizontally-partitioned batch data, each for a different overlord server


My application has horizontally partitioned data. For each time stamp, there are multiple rows (each row has multiple columns). The rows of data are partitioned into different batch files according to our business logic.

Question is, can I setup multiple overlord servers, and each ingest a different batch data?

We have lots of data for each time stamp, that’s why we have this horizontally partitioned data to make it distributed.

With this use case in mind, what would you recommend?