Databases

Question 1

Would you use an RDS database for OLTP or OLAP?

Answer 1

For Online Transaction Processing (OLTP) workloads.
It’s not suitable for analysing large amounts of data.

Question 2

What do OLTP and OLAP stand for?

Answer 2

Online Transaction Processing
Online Analytical Processing

Question 3

What’s the difference between OLTP and OLAP?

Answer 3

OLTP:
- Real time data
- Data processing
- Large numbers of small transactions.

OLAP:
- Historical data
- Data analysis (large amounts of data)
- Complex queries

Question 4

What does RDS Multi-AZ do?

Answer 4

It creates an exact copy of your production database in another AZ.
(Writes are automatically synchronised to the standby database.)

Question 5

Which RDS type is always Multi-AZ?

Answer 5

Aurora

Question 6

RDS Multi-AZ is for:
A: Improving performance
or
B: Disaster recovery
?

Answer 6

B: Disaster recovery

You can’t connect to the standby while the primary database is active.

Question 7

What would you typically use for OLAP workloads?

Answer 7

Redshift

Question 8

TRUE or FALSE?
RDS read replicas must be in a different AZ to the primary

Answer 8

FALSE:

They don’t have to be but they can be.

They can also be in a completely different region.

Question 9

With RDS Multi-AZ, what happens if the primary fails?

Answer 9

It changes the DNS endpoint to point to another IP address (of the secondary).
i.e. it automatically fails over

Question 10

TRUE or FALSE?
Read replicas can be promoted to be their own databases.

Answer 10

TRUE
However this breaks replication.

Question 11

TRUE or FALSE?
Automatic backups must be enabled in order to deploy an RDS read replica.

Answer 11

TRUE
The replication process involves using the most recent backup to set up the read replica.

Question 12

How many read replicas can you have to each RDS instance (except Aurora)?

Answer 12

5

Question 13

When would you use a read replica? (2 things)

Answer 13

To get better read performance.

To take the load off the primary database for read-only workloads.

Question 14

What databases is Aurora compatible with?

Answer 14

MySQL and PostgresSQL

Question 15

How many copies of your data do you get with Aurora (minimum)?

Answer 15

6
2 copies in each AZ over 3 AZs (minimum)

Question 16

What are the 3 types of Aurora replicas? How many read replicas can you get with each?

Answer 16

Aurora (15), MySQL (5), and PostgresSQL (5)

Question 17

TRUE or FALSE?
Backups are always enabled on Aurora

Answer 17

TRUE

Question 18

TRUE or FALSE?
Aurora backups impact database performance

Answer 18

FALSE

Question 19

What would you use if you need the performance of Aurora but you have spiky workloads?

Answer 19

Aurora Serverless

Question 20

TRUE or FALSE?
You can share Aurora snapshots with other AWS accounts.

Answer 20

TRUE

Question 21

Whats the difference between eventually consistent read and strongly consistent reads?

Answer 21

Eventually - consistency reached within 1 second
Strongly - result will contain all writes prior to the read

Question 22

What is DAX?

Answer 22

An in-memory caching service for DynamoDB.

It improves performance as the application doesn’t need to try the cache first, it queries DAX directly.

Question 23

What performance improvement can you get with DAX compared to using DynamoDB alone?

Answer 23

Up to 10x

Question 24

What does ACID stand for?

Answer 24

Atomic
Consistent
Isolated
Durable

Question 25

What is the ACID methodology?

Answer 25

A set of properties that guarantee reliability and consistency in database transactions.

It basically means all or nothing. Either a transaction succeeds or it doesn’t.

Question 26

What does DynamoDB Transactions provide?

Answer 26

ACID
(All or nothing approach)

Question 27

What are some use cases for DynamoDB transactions?

Answer 27

Financial transactions
Managing orders
Multi-player game engines
Coordinating actions across distributed components and services

Question 28

Exam tip: If you see a scenario question that mentions ACID requirements, think of…

Answer 28

DynamoDB Transactions

Question 29

In DynamoDB, read and write capacity units are allocated based on…

Answer 29

partition

Question 30

TRUE or FALSE?
In DynamoDB, Point in Time Recovery is enabled by default.

Answer 30

FALSE

Question 31

To use DynamoDB Global tables, make sure that ____ is enabled.

Answer 31

Streams

Question 32

DynamoDB is integrated with _____ so that you can createtriggers—pieces of code that automatically respond to events in DynamoDB Streams.

Answer 32

AWS Lambda

Question 33

In DynamoDB, it’s best practice is to choose partition keys with high cardinality.

Answer 33

TRUE. This way there are many unique values.

Question 34

TRUE or FALSE?
Amazon Redshift is serverless.

Answer 34

FALSE
It uses EC2 instances.