Ship with a local DB as a possible source/destination

[fleupold]

This is probably out of scope for milestone 3, but given we have a use case for it, I'd rate it more important than S3 buckets (which was already labelled for a yet to negotiate milestone 4).

We would like to be able to do the following

1. Run a query on some DB1 and mirror it into a temporary table/storage
2. Run another query on some DB2 and mirror it into a temporary table/storage
3. Combine the two results and upload the final table to Dune

Currently this is not possible. Potentially using csv files (which I assume would be part of #59) a strict "append" combination would be possible.

However, to be maximally flexible I wonder if the container couldn't ship with a built in local sqlite or postgres DB that could be used as a data source and sink?

Then, we could order jobs accordingly to make sure the local table is populated before it gets synced to Dune.