and it should work without any fault tolerance for now. Node when it gets that query it shoud ask master for a list of nodes that have data from this table. Then it should query all this nodes for data using the same sql. Next it should concatenate data with distinct on ID. Caution: ALL DATA IN OUR DB MUST HAVE ID COLUMN WITH UNIQUE AS INT - it is for now, later we can add additional column for our ids. We will also add some additional columns for timestamp (version) for example.
User should be able to query SELECT * FROM