How to choose a method for storing online store data?

Question

NoSQL

Good day to all.
I need to make an online store for shoes and accessories.
I want to use monga because: 1. it’s
fashionable (where would it be without it?!)
2. I want to try it exactly (to practice so to speak)
3. I’ll write the server part in Go, and it’s almost impossible to work with complex sql queries in Go ( problems with query result checks - joins are not supported by the current driver, etc.).
In this regard, I want to figure out in order not to step on a rake: how best to organize the structure of storing data about a product/categories?
A few words about how the catalog looks now.
Product data is stored in 1s(UT). They will be uploaded to a json file. Next, you need to parse this file and drive it into monga.
The structure of the nomenclature is as follows:
Brand 1
Collection1
...Products...
Collection2
...Products...
Brand 2
Collection1
...Products...
Collection2
...Products...
Product characteristics should, in theory, be the same (price, color, material), but not all. There are such things as different sizes from different suppliers (small sizes / under the domestic leg / European sizes, etc.).
There are ideas (I saw somewhere) that you need to organize a document with categories:
{main category, child category}
And assign a correspondence, main category and child category to each product. Is this how it should be done?
I will be glad for any information.
Thanks in advance!)

Answer 1

In my opinion, the storage architecture is not very correct.
Make two collections. One for Categories and one for Products.
Add the following fields to the Product: Brand, Category [array].
According to the characteristics of the product - they, alas, are not the same for me, so I store them in the Product as an array of objects

{ name: Сапог сложный
  sku: smart-shoozze-$Цвет$-$Размер$
....
Характеристики: [
{name: Цвет, value: Зеленый, type: checkboх, sku: GREEN},
{name: Цвет, value: Красный, type: checkboх, sku: RED}
]
....
}

To display filters - do aggregation on this field.
If perishing essentially - still there is a collection Remains. There is also a collection of Media Files.
If it's very deep, it's done on elastic, but there shouldn't be any problems with monga.

Answer 2

1. fashionable (where without it ?!)

Monga advantage only in clusters
https://www.youtube.com/watch?v=SNzOZKvFZ68
JOINs in Mongo are not normally supported at all.
This is not a relational DBMS.
Try an ORM reform rather than the standard Go driver.
In a documentary DBMS, which Monga is, entities are not divided into tables (normal form) as in relational ones.
Keep everything together.