Misc

Question 1

What is Elastic Beanstalk?

Answer 1

It pre-provisions resources for a given result.

Question 2

What is the main difference between Elastic Beanstalk and CloudFormation?

Answer 2

Beanstalk is more automated

Question 3

What is CloudFormation?

Answer 3

Infrastructure as code

Automating resource provisioning

Question 4

What is a CloudFormation stack?

Answer 4

One or more CloudFormation templates that will provision resources

Question 5

Adding hardware to an EC2 instance is an example of ____ ____ vs adding more EC2 instances to a group, which is ____ ____

Answer 5

Scaling up

Scaling out

Question 6

What are the four methods of scaling out (horizontally)?

Answer 6

Stateless Applications

Stateless Components

Stateful Components

Distributed Processing

Question 7

How do stateless appplications support horizontal scaling and what is an example of a stateless application?

Answer 7

Since session information is not stored on the application hardware so scaling out does not rely on any state data. Lambda is an example of this

Question 8

When distributing workload to multiple nodes, what are the two models and what AWS services support each?

Answer 8

Push Model - ELB (NLB, ALB), Route53 round robin

Pull Model - SQS, Kinesis

Question 9

How can horizontal scaling be supported with stateless components?

Answer 9

Cookies in users browser

DynamoDB

Question 10

What is the relationship between horizontal scaling and stateful components?

Answer 10

Most real-world apps require some state that can’t be kept as cookies.

Implement session affinity with the sticky session feature of an Application Load Balancer

Question 11

How can distributed processing support horizontal scaling?

Answer 11

For large very large datasets

Services like Elastic Map Reduce can break up data and distribute it

AWS Glue

AWS Batch

Question 12

What are the overall Cloud Design Principles?

Answer 12

Scalability

Disposable Resources over Fixed Servers

Automation

Loose Coupling

Services, Not Servers

Databases

Managing Increasing Volumes of Data

Removing Single Points of Failure

Optimize for Cost

Caching

Security

Question 13

How do we architect for disposable resources instead of fixed servers?

Answer 13

Bootstrapping (EC2 launch scripts)

Golden Images (Prebuilt AMIs)

Hybrid of two

Infrastructure as Code (CloudFormation)

Question 14

What are the two services in AWS that support Infrastructure as Code?

Answer 14

Elastic Beanstalk

CloudFormation

Question 15

How can the AWS Automation Principle be supported?

Answer 15

Loose Coupling

Question 16

What are the components of Loose Coupling?

Answer 16

Well defined interfaces

Service Discovery (no hardcoded IPs)

Async Integration (queues)

Graceful Failure

Question 17

How is the principle “Services not Servers” implemented?

Answer 17

Using managed services (Paypal)

Serverless Architectures (Lambda, API Gateway, DynamoDB)

Question 18

What are the keys to Removing Single Points of Failure?

Answer 18

Introduce redundancy

Detect Failure (health checks)

Durable Data Storage (S3, Dynamo, not EC2)

Multiple AZ

Shard data (break it up)

Question 19

What are the components of the Security Principle?

Answer 19

Utilize AWS features

Offload respnsibility to AWS

Reduce Priveleged Access

Security as Code

Audit in Real Time