What request will faster work?

Let's say you need to make a method that will get the old associated with 3-5 models. How to do it better, and most importantly what will work faster:
1. All queries in the database to make possible parallel or in place?
2. To usnot aggregation and $lookup?
With the aggregation never worked, but cheto it seems to me that such variant will work faster but will be harder to ship database, although is not a fact.
June 14th 19 at 18:05
1 answer
June 14th 19 at 18:07
1. well in General if the hardware allows, the parallel will be faster to work with.
solid state drives in multiple threads allow for faster data read.
2. built-in mechanisms to work, the database is usually faster to work than to build this logic yourself, Mongo allows you to store a "reference" data (which if you request for yourself are loaded from other tables).
3. Plus the Mongo that you have all the answers just a push in the collection. For example, you have one database with some data, and another database with these data but organized on a different principle, which for example can find all the answers, all the relationships and so on, try to implement another base, with pre-aggregation of all required data.

Find more questions by tags MongoDBNode.js