NoSQL and MongoDB

Question 1

Problem with Relational Databases

Answer 1

• Do not scale well to large clusters

Question 2

Sharding

Answer 2

• Divides up the rows of a collection of databases across multiple servers

Question 3

Characteristics of NoSQL Databases

Answer 3

• Don’t use the relational model
• Open-source
• Designed for clusters (sharding, fragmentation, replication)
• Schema-less or flexible

Question 4

Three Main Classes of NoSQL Databases

Answer 4

• Key-value stores
• Document databases
• Column-family (big-table) stores

Question 5

Key-Value Stores

Answer 5

• Queries are limited to the key

Question 6

Document Stores

Answer 6

• Values are not opaque
• Queries can be based on the key or the contents
• Support for sharding and replication

Question 7

Aggregate Orientation

Answer 7

• Group together data that belongs together
• Get related data without accessing many different nodes

Question 8

Advantages of Schemalessness

Answer 8

• Data types can evolve over time
• Easier to handle nonuniform data

Question 9

Disadvantages of Schemalessness

Answer 9

• Can’t be enforced by DBMS
• DBMS can’t make queries more efficient

Question 10

MongoDB Equivalent of “Table”

Answer 10

• Collection

Question 11

MongoDB Equivalent of “Row”

Answer 11

• Document

Question 12

MongoDB Equivalent of “Attributes”

Answer 12

• Fields

Question 13

MongoDB Equivalent of “Primary Key”

Answer 13

• _id

Question 14

Capturing Relationships in MongoDB

Answer 14

• Store references to other documents using their _id values
• Embed documents within other documents
• Use references if embedding could lead to significant growth in the size of the document over time

Question 15

Limitations of MongoDB

Answer 15

• A query can only access a single collection
• Atomicity is only provided for operations on a single document

Question 16

MongoDB Query Selection

Answer 16

• db.collection.find(<selection>, <projection>)</projection></selection>

Question 17

MongoDB Single-Purpose Aggregation

Answer 17

• db.collection.count(<selection>)</selection>
• db.collection.distinct(<field>, <selection>)</selection></field>

Question 18

MongoDB Aggregation Pipeline

Answer 18

• db.collection.aggregate({<pipeline-op1>: <pipeline-expression1>})</pipeline-expression1></pipeline-op1>

Question 19

MongoDB Pipeline Operations

Answer 19

• $match
• $group
• $project
• $unwind
• $limit
• $sort