1
00:00:02,340 --> 00:00:08,520
So we had a look at the general data structure but always only for one collection and the documents

2
00:00:08,520 --> 00:00:12,280
in that the one collection. Typically you have multiple collections though,

3
00:00:12,480 --> 00:00:16,470
you might have users and products and orders.

4
00:00:16,470 --> 00:00:23,490
So if you have multiple collections that still are related of course or where the documents in these relations

5
00:00:23,490 --> 00:00:24,580
are related,

6
00:00:24,690 --> 00:00:29,570
you obviously have to think about how do you store related data,

7
00:00:29,760 --> 00:00:37,650
do you use embedded documents because that of course is one way of reflecting a relation, if a user

8
00:00:37,710 --> 00:00:41,640
has an address and that address consists of a street and a city,

9
00:00:41,730 --> 00:00:45,680
then you might have a nested address document in the user document

10
00:00:45,870 --> 00:00:48,590
or do you work with references?

11
00:00:48,630 --> 00:00:49,930
Here are two examples,

12
00:00:50,130 --> 00:00:54,580
as I just said, we might have a nested document and an embedded document,

13
00:00:54,600 --> 00:00:55,870
here are my customers

14
00:00:56,040 --> 00:01:02,160
and by the way of course in json, the field names like username, age and so on would be enclosed in double quotation

15
00:01:02,160 --> 00:01:02,790
marks,

16
00:01:02,790 --> 00:01:05,510
I'm just leaving that out here to save some space,

17
00:01:05,520 --> 00:01:09,460
so this is not a real json code but it is structure of the data.

18
00:01:09,690 --> 00:01:15,570
And here, my customer has an address and that address could totally be stored in a new collection

19
00:01:15,570 --> 00:01:22,170
but we can also embed it into our user or into our customer here because it's strongly related to that

20
00:01:22,170 --> 00:01:28,590
customer and we might want to fetch both at the same time or at least have an easy way of fetching both

21
00:01:28,710 --> 00:01:30,320
if we need to.

22
00:01:30,330 --> 00:01:33,180
The alternative is that you use references,

23
00:01:33,180 --> 00:01:40,530
let's say we still have customers and each customer has a list of his favorite books and we could absolutely

24
00:01:40,530 --> 00:01:43,390
go with nested or embedded data.

25
00:01:43,410 --> 00:01:49,860
So here we have an array with embedded documents and each document is shortened here but each document

26
00:01:49,860 --> 00:01:54,840
could be a book with a title, a price, an author and so on.

27
00:01:54,840 --> 00:01:59,790
Now the problem here is that we will have a lot of duplicate data because different customers can have

28
00:01:59,790 --> 00:02:01,800
the same favorite books

29
00:02:01,800 --> 00:02:07,000
and if we then change something about that favorite book, we'll have to change it on all customers,

30
00:02:07,020 --> 00:02:12,970
so instead here, we might go for references. We might have a customers collection and the books collection

31
00:02:13,500 --> 00:02:20,730
and in the customer, we still have that array of favorite books but there, we only store the IDs of the

32
00:02:20,730 --> 00:02:22,070
favorite books.

33
00:02:22,110 --> 00:02:26,220
That means that when we fetch the customers and we also want to fetch the favorite books for

34
00:02:26,220 --> 00:02:29,760
the customer, we have to run two queries, that is correct

35
00:02:29,970 --> 00:02:35,140
but on the other hand if we change a book, we don't have to change it on 10 different customers but only

36
00:02:35,140 --> 00:02:38,220
in the books collection and the ID will never change,

37
00:02:38,220 --> 00:02:43,590
so we never have to change or touch the favorite books array in the customers document,

38
00:02:43,860 --> 00:02:47,070
so we have that relation through a reference.

39
00:02:47,070 --> 00:02:53,370
Now you got these two different approaches and now the obvious question is when should you use which.

40
00:02:53,910 --> 00:02:56,190
Time to have a look at some examples.