RethinkDB: Getting random documents for each category - database-performance

Question

I have a table with ~8000 quiz questions.
They are divided in 25 categories.
Each category has a attribute max_questions which tells me how many questions i have to pick randomly to generate a quiz.
e.g
Category 1 -> 2 questions
Category 2 -> 3 questions
Category 3 -> 1 question
I came up with a solution, but i takes approx. 2 seconds to perform.
r.table('categories').pluck('id', 'max_questions').orderBy('id').run(conn, function(err, cursor) {
if(err) return next(new Error(err.msg));
cursor.toArray(function(err, categories) {
if(err) return next(new Error(err.msg));
async.concat(categories, function(category, callback) {
r.table('questions').filter({category_id: category.id }).sample(category.max_questions).run(conn, callback);
}, function(err, questions) {
if(err) return next(new Error(err.msg));
res.json(questions);
});
});
});
Is there a faster way to retrieve the questions with RethinkDB? Making 25 requests and calling 25 times .sample() for one quiz doesn't sound good to me.
I really appreciate your help!

Answer 1

It'll be much faster if you do it all in one query rather than making multiple requests to the database. Here's a single query that's more or less equivalent to what you wrote:
categories.map(function (doc) {
return doc.merge(
{"questions":
questions
.filter({category_id:doc("id")})
.sample(doc("max_questions"))
.coerceTo("ARRAY")})
})
Notice I've bound the tables to variables here so categories is bound to r.table("categories").