This post was actually written on my blog at www.easyontheweb.com <------

If you’re a beginner you might have or might not have heard of indexing in databases. This is not something specific for MongoDB but today I’ll explain some basics of MongoDB indexes for beginners.

What is Indexing?
What indexing does is sort our mongoDB collection in a particular order based on the value of one field(or more than one field). Assume that I have a collection called customers and I have kept customerName as the field for indexing, then what MongoDB does is that it’ll create a list of all the names in an alphabetical order in the document.

The list will be just the names and each item in the list will contain a pointer to the real document in the collection. What this does is the next time you run a find query with the customerName as the filter parameter, Mongo will directly look into that list and easily and quickly find your required document(s).

Pros and cons of indexing
The biggest advantage of indexing is that it speeds up your find, update and delete queries. Quite naturally because it is easier to search for the elements based on the indexed field.

The disadvantages of indexing is that 1. It takes up memory (obviously). 2.It slows down write queries.

The write queries will obviously be slowed down because every time you make a write query you need to update the indexed field list in the collection as well and sort it again based on that field.

Creating Indexes in Mongo
Let us see how can you create indexes in Mongo.

First of all there are three different kinds of indexes that you should probably know of :-
1.Single index -> Where you sort a collection based on just a single field value.
2.Compound index -> Where you sort a collection first based on a single value and then for the values that have the same first value, you sort the list according to a second field value that you have provided.
3.Partial index -> When you sort the collection based on a field value but only in a particular range (we’ll see how later).

Single Index

1
2
3
4
db.collection.createIndex({fieldName: 1});

example

db.customers.createIndex({customerName: 1})
The 1 represents ascending order of list sorting.

Compound Index

1
db.customers.createIndex({customerName: 1, age: 1});
What this does is create an index based on the customer name first and then the age (if two or more customer names are the same). Note that for compound indexes, you can use them to index for the leftmost field or all the fields moving left to right. They will speed up your queries for both. But this indexing will not work if you think of querying only over age.

Partial Index

Assume we run a particular query more often than not. For example, we often search for customers with age less than 19 and we only want to create an index on the age field for documents where is less than 19, what this will do is not slow down our insertions for documents where age is greater than 19.

1
db.customers.createIndex({age: 1, {partialExpression: {age: {$lt: 19}}}});

I hope you got the basics of indexing and when to use indexes and which fields to use indexes on -> basically create indexes for the field you run the most queries on. Note that it is a bad idea to create an index for each field as it slows down your insertions by a lot and takes up a lot of memory.