Data and Analytics

Question 1

What is Amazon Athena?

Answer 1

A server-less query service used to analyse data stored in Amazon S3 with SQL queries

Question 2

What are federated queries?

Answer 2

Queries that can be run across multiple data sources than just what is in S3, such as relational, non-relational, object and custom data sources

Question 3

What is 1 method that can be used to increase the performance of Athena?

Answer 3

Partitioning
Using columnar data
Use larger files as these are easier to scan and retrieve for Athena

Question 4

What is Amazon Redshift used for?

Answer 4

Data warehousing and analytics

Question 5

Is Redshift columnar or row-based?

Answer 5

Columnar

Question 6

What engine is Redshift based on?

Answer 6

PostgreSQL

Question 7

What are the two snapshot modes of Redshift and what are the differences?

Answer 7

Automated and manual.
With automated, the snapshot is retained for a period that the user sets, whereas with manual the snapshot is kept until it is deleted.

Question 8

What are the two node types within a Redshift cluster?

Answer 8

Leader and compute

Question 9

What is Redshift Spectrum?

Answer 9

A service that allows the user to query data that is already in S3 without having to load it

Question 10

What is the principal benefit of Redshift spectrum?

Answer 10

It allows the user to leverage a lot more computing power than they actually have provisioned and for the avoidance of having to actually load the S3 data

Question 11

What is OpenSearch?

Answer 11

A service that allows the user to search any field, including partial matches, of a database

Question 12

What is EMR?

Answer 12

Elastic Map Reduce - a service that allows the user to create Hadoop clusters for big data analytics

Question 13

How does EMR scale?

Answer 13

Automatically, through the provisioning of additional clusters

Question 14

What are the node types within an EMR cluster?

Answer 14

Master, core and task.
Master nodes manage the cluster and co-ordinate the other nodes. There is only 1 in a cluster.
Core nodes run tasks and store data.
Tasks nodes are optional and just run tasks but don’t store data.

Question 15

What service would be used to make ML-powered interactive dashboards?

Answer 15

QuickSight

Question 16

When can QuickSight not use SPICE in-memory computation?

Answer 16

When it is attached to another database

Question 17

What granularity of security can you set in QuickSight?

Answer 17

Column-level security

Question 18

What does ETL stand for?

Answer 18

Extract, transform and load

Question 19

What is Glue?

Answer 19

A managed and server-less ETL service for analytics

Question 20

What are Glue Crawlers?

Answer 20

Scripts that crawl databases or data and write metadata to Glue Data Catalog, e.g. the type of data and its format

Question 21

What are Glue Job Bookmarks?

Answer 21

Bookmarks that show where a job was up to, preventing the re-processing of data

Question 22

What is Glue DataBrew?

Answer 22

A service that cleans and normalises data for analytics and ML without having to write code - many pre-written transformations

Question 23

What is Lake Formation and data lakes?

Answer 23

A data lake is a central place to keep all data of different types for analytics purposes.
Lake Formation is an AWS service that simplifies the process of creating a data lake through the automation of many complex processes.

Question 24

What level of granularity does Lake Formation have in terms of security?

Answer 24

Row/column level

Question 25

What service would be used for real-time analytics using SQL?

Answer 25

Kinesis Data Analytics for SQL

Question 26

Where can Kinesis Data Analytics for SQL read from?

Answer 26

Kinesis Data Streams and Kinesis Data Firehose

Question 27

What is a benefit of using Kinesis Data Analytics for Apache Flink over for SQL?

Answer 27

Flink is more powerful with more advanced querying that just using SQL

Question 28

What is Amazon MSK?

Answer 28

Managed Streaming for Apache Kafka ( a data streaming alternative to Kinesis) - fully managed Kafka on AWS

Question 29

Are Kinesis Data Streams’ streams encrypted?

Answer 29

Yes, in-flight using TLS