Saturday 21 February 2015

Designing one-to-one mapping in mongoDB


In one-to-one relation, each item corresponds to exactly one item.

Example
Employee-Resume
student-marksSheet
customer-transaction
person-passport

One employee has one resume associated with it. One resume corresponds to one employee.

One student has one marks sheet and one marks sheet associated with one student.

Let’s try to model this one to one mapping for student-markssheet.

There are three ways we can model this relation
1. Embed marks_sheet document in student document
2. Embed student document in marks_sheet document
3. Define student and marks_sheet documents separately.

Embed markssheet document in student document
First way is we can embed marks_sheet document in student document.

{
        "_id" : 512,
        "firstName" : "Hari Krishna",
        "lastName" : "Gurram",
        "marks_sheet" : {
                "mathematics" : 98,
                "physics" : 85,
                "chemistry" : 79
        }
}


Embed student document in marks_sheet document
Second way is we can embed student document in marks_sheet document.


{
        "_id" : 1,
        "mathematics" : 98,
        "physics" : 85,
        "chemistry" : 79,
        "student" : {
                "id" : 512,
                "firstName" : "Hari Krishna",
                "lastName" : "Gurram"
        }
}


Define student and marks_sheet collections separately.
student document contains a filed “marks_sheet_id” which is used to identify “marks_sheet” document.

{
        "_id" : 512,
        "marks_sheet_id" : 1,
        "firstName" : "Hari Krishna",
        "lastName" : "Gurram"
}


marks_sheet document contains a field “student_id”, which is used to identify “student” document.

{
        "_id" : 1,
        "student_id" : 512,
        "mathematics" : 98,
        "physics" : 85,
        "chemistry" : 79
}


Ok cool, so I had three ways, which one I have to use  

It is totally depends on type of data you are going to model. Below are the key points you need to consider while choosing approach.



1.   Frequency of access
If you are going to access student documents more frequently and marks_sheet document less frequently, then define student and marks_sheet documents separately. It is because if you go for embedding documents approach, you are unnecessarily loading marks_sheet document, whenever you loads student document.

2.   Size of document
If the embedded document size is very large, and even of you want to update outer document (not embedded), you have to load entire document. So in this case separate documents.

 
Prevoius                                                 Next                                                 Home

No comments:

Post a Comment