MongoDB Cheatsheet

1955



1. About MongoDB

MongoDB is a cross platform document oriented NoSQL database.

2. Installation

On Mac

brew install mongodb

On Ubuntu (18.04)

sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv 9DA31620334BD75D9DCB49F368818C72E52529D4
echo "deb [ arch=amd64 ] https://repo.mongodb.org/apt/ubuntu bionic/mongodb-org/4.0 multiverse" | sudo tee /etc/apt/sources.list.d/mongodb-org-4.0.list
sudo apt-get update
sudo apt-get install -y mongodb-org

For other platforms you can get details here https://docs.mongodb.com/manual/administration/install-community/

3. Mongo Shell

Mongo Shell is easiest way to experiment with MongoDB in the beginning.
You can login to shell by running command mongo

4. Mongo Commands

  • list all databases
show dbs
  • create/ login to a database
use <database_name>
* create collection
```sh
db.createCollection('<collection_name>')

5. Mongo Queries

Lets take an use case, using which we can go over all MongoDB commands.
Lets build a 'Car Information' application which stores all cars with its specifications.

Inserting documents

  • db.collection.insert()

using insert you can either insert one document or array of documents

db.cars.insert({  "name" : "Volvo xc 60",  "type" : "SUV" });
db.cars.insert(
[
  { "name" : "Volvo xc 60",  "type" : "SUV" },
  { "name" : "Volvo xc 90",  "type" : "SUV" },
  { "name" : "Volvo   S90",  "type" : "SUV" }
]
);

  • db.collection.insertOne()

Inserts one document

db.cars.insert({  "name" : "Volvo xc 60",  "type" : "SUV" });

  • db.collection.insertMany

Inserts multiple documents

db.cars.insertMany(
  [
    { "name" : "Volvo xc 60",  "type" : "SUV" },
  ]
);

  • db.collection.save()

Upsert a document. If you provide _id which is already exist then it updates the document.

db.cars.save( { _id: 3, "name" : "Volvo xc 60",  "type" : "SUV"} );

Updating documents

  • db.collection.update()

Updates one or more than one document(s) in collection based on matching document and based on multi option

Syntax:

db.cars.update(
 <query>,
 <update>,
 <options>
);

Example:

db.cars.update(
  { type: "SUV" },
  { $set: { size : 5} },
  { multi: true}
)

  • db.collection.updateOne()

Updates a single document based on matching query

  • db.collection.updateMany()

Updates multiple documents based on query

  • db.collection.replaceOne()

Replaces entire content of document except _id field

  • db.collection.findOneAndUpdate(<filter>, <update>, <options>)

Updates a single document. Following are some of the options
upsert : When true inserts a document.
returnNewDocument: when true returns new document instead original document. When upsert is true and returnNewDocument is false then null will be returned in case of new document insertion.

Reading Documents

  • db.collection.findOne()

Returns one document matching the given query

db.collection.findOne({ _id : 123}); # finding a document with _id
db.collection.findOne({ type : "SUV"}); # return one SUV

findOne with projection (limiting the fields to return)

db.collection.findOne({ type : "SUV"}, {name: 1}); # returns _id & name fields only

  • db.collection.find()

Returns a cursor with selected documents

 db.collection.find({ type : "SUV"}); # returns all SUVs

Deleting Documents

  • db.collection.deleteOne(<filter>, <options>)

    Deletes a Single document from collection

  • db.collection.deleteMany(<filter>, <options>)

    Deletes all documents with matching filter

6. Read Concern

We can control the consistency and isolation properties of data reads ( from replica sets & replica set shards) using readConcern option

Read Concern Levels:

  1. local: data returns from the instace without guaranteing that its been written to majority of replica members

Default for reads against primary & reads against secondaries
2. available: similar to local, gives lowest latancy for sharded collections.
Default for reads against secondaries
3. majority: only if the data acknowledged by a majority of the replica set members.
4. linearizable: return data after all successful majority-acknowledged writes
5. snapshot: Only available for transactions on multi documents.

7. Write Concern

Using Write Concern we can set the level of acknowledgment for a given write operation.
This will be sent to mongod and mongos (in case of sharded collections)

Following are the fields that specify write concern

{
 w : <number>,
 j : <boolean>,
 wtimeout: <milli seconds>
}

w lavlues:
0: No acknowledgement requested.
1: This requests for an acknowledgement that is propegaded to standalone mongod or primary in replica set
Note: This is the default write concern value for MongoDB
Greater than 1: Requests acknowledgement from primary & given number -1 of secondaries

j option:
This requests acknowledgement on whether or not the write operation written onto on-disk journal files

wtimeout:
Operations will return error after specified limit, this prevent the client waiting indefinitely.
On failure MongoDB does not rollback the data, data may eventually get stored

8. Indexes

_id Index (Default index):

We can't drop this index

1. Single Field:
2. Multikey Field:
3. Geospatial Index:
4. Text Indexes:

9. Aggregations

1. Aggregation Pipeline
2. Map-Reduce
3. Single Purpose

10. MongoDB Shell Help

help