I have the following simple query
select row_to_json(d) from documents d
The output of this goes to script that expects new-line-delimited stream of JSON objects.
But as-is, ti looks like the server’s memory fills-up before ti starts emitting results.
Any ideas how I could bypass this?