Queries endpoint should be able to respond a query of 1M rows

[ferblape]

The queries endpoint should be able to respond to user queries that fetch a considerable amount of rows.

To define scope, I'd try to reach the 1M row milestone (and plan a different issue for 10M). I'd tackle this in an iterative way:

• objective 1: 10,000 rows
• objective 2: 100,000 rows
• objective 3: 500,000 rows
• objective 4: 1,000,000 rows

Looks like using streaming could be a good way to implement this, and modern browsers and JS libraries support it.

It's important to understand once we reach the limit, what is causing that limit and what steps we need to follow to reach the next mileston of 10M.