Databases

Question 1

Is RDS OLTP or OLAP?

Answer 1

RDS is OLTP as it’s row-based and not made for data warehousing.

Question 2

What OLTP Databases are available on AWS?

Answer 2

MySQL
MariaDB
Aurora
PostgreSQL
Oracle
SQL Server

Question 3

What OLAP service is there on AWS?

Answer 3

Redshift (dataware housing)

Question 4

What’s the difference between OLAP and OLTP?

Answer 4

OLAP is for analytics, OLTP is for transaction processing.

OLAP uses columnar architecture to achieve high IO performance (same data type in a data column) as well as very good compression. Queries can span multiple tables and process/join multiple sources of data which can be very intensive.

OLTP uses row-based architecture which can be easily queried without spanning too many databases.

Question 5

What is AWS Redshift?

Answer 5

Redshift is a data warehousing service for AWS. It has fast performance and high storage capacity. It is a columnar type architecture with multi-parallel processing and advanced data compression.

Question 6

Is AWS Redshift encrypted?

Answer 6

At rest encryption: AES-256

In transit encryption: SSL/TLS

Question 7

What is Multi-AZ in an RDS Instance?

Answer 7

Multi-AZ is used to synchronize your database writes to a copy in another AZ so that it can act as a failover replica in the case your primary database fails.

Used for High Availability

Question 8

What is Read-Replicas in an RDS instance?

Answer 8

Read-Replicas are read-only replicas of your primary database that you can use to gain performance/separate reads from your primary database.

This is used to gain a performance advantage.

Question 9

What are the available database services in AWS?

Answer 9

OLAP: Redshift
OLTP: RDS – MySQL, SQL Server, PostgreSQL, Oracle, Aurora, MariaDB

In-Memory: Elasticache (redis, memcached)

No-SQL: DynamoDB

Question 10

How does Aurora Scaling/Failover work?

Answer 10

Two copies of your database in each AZ. Each Aurora instance has a minimum of three AZs (6 copies of physical storage media)

Designed to transparently handle the loss of 2 copies of data without affecting database write availability and up to three copies without affecting read availability

Self-healing, data blocks and disks are continuously scanned for errors and repaired automatically.

Question 11

How many copies of data can Aurora handle?

Answer 11

Aurora is able to handle up to 2 copies of lost data without affecting write operations and up to 3 copies of data without affecting read availability

Question 12

What stops an Aurora instance from suffering data corruption/drive corruption?

Answer 12

Self-healing. Aurora constantly scans for errors and repairs them automatically

Question 13

Difference between scaling DynamoDB vs scaling RDS?

Answer 13

DynamoDB can scale with “push button scaling” aka you click a button and you can increase the performance (WCU,RCU) units and size. No downtime required.

With an RDS instance you have to use read-replicas to scale out and eventually hit a limit or use a bigger instance type.

Question 14

What are the storage consistencies of Dynamo DB?

Answer 14

Eventual Read Consistency (default) : Can take 1 second to reach consistency after a write. (.5 WCU)

Strongly Consistent Reads: Less than a second to reach consistency after a write (1 WCU)

Question 15

How many data centers are DynamoDB instances spread across?

Answer 15

DynamoDB instances are spread across 3 geographically distinct data centers

Question 16

What type of storage are DynamoDB instances stored on?

Answer 16

SSD Storage

Question 17

What types of Redshift Configurations are there?

Answer 17

Single node (164GB): Used for small/med business

Multi-Node
Leader Node: Manages client connections and receives queries

Compute Nodes (up to 128 nodes): store and perform queries and computations

Question 18

What is Elasticache?

Answer 18

A web service that caches critical data that requires high performance reads. Increases performance by storing data in-memory instead of on disk.

Question 19

What are the two Elasticache caching engines?

Answer 19

Memcached

redis

Question 20

When would you use Elasticache?

Answer 20

When you need to offload reads or higher read performance from data.

Question 21

When would you use Redshift?

Answer 21

When you’ve got queries that span too many tables, databases, taking up tons of I/O

Question 22

How many copies of your data are stored by default with Aurora? (physically)

Answer 22

6 copies (2 copies across 3 different AZs)

Question 23

What is Amazon Athena?

Answer 23

Athena is a query service that makes it easy to analyze data in S3 using standard SQL commands.

Supported data formats: JSON, Apache Parquet, Apache ORC

Question 24

What happens to the I/O of a single-AZ RDS instance during a DB snapshot or backup?

Answer 24

You may experience I/O being suspended briefly as the backup process initializes (under a few seconds). You may also experience higher latency for a period.

Question 25

Default MySQL port number?

Answer 25

3306

Question 26

With new RDS DB instances are automated backups enabled by default?

Answer 26

Yes, automated backups are enabled by default

Question 27

What is the maximum retention interval for RDS backups?

Answer 27

35 days

Question 28

Do you have to pay a transfer charge when replicating data from your primary RDS instance to your secondary RDS instance?

Answer 28

Nope, there is no charge associated with the action.

Question 29

Are you charged if you have an RDS database that isn’t being used?

Answer 29

Yes, RDS databases are charged based on running time