Elastic Load Balancing & Auto Scaling Groups Section

Question 1

Scalability

Answer 1

Abillity to handle a large load by making the hardware stronger (scale up) or by adding nodes (scale out).

There are two types of Scalability:
1. Vertical
2. Horizontal

(Page 113 & 118)

Question 2

Vertical Scalability?

Answer 2

Increase de SIZE OF THE INSTANCE.
It’s limited by Hardware.
It’s not common in distributed systems

(Page 114)

Question 3

Horizontal Scalability?

Answer 3

Increase THE NUMBER OF INSTANCES.
Implies Distributed Systems.
Common on Web and modern applications.

(Page 115)

Question 4

High Availability?

Answer 4

Capacity to survive disasters.
To have a High UpTime of working.
In AWS, means to run your App on at least two different AZ.
Works well with Horizontal Scaling.

In Nute Shells: Run the same instance across multiple AZ.

(Page 117)

Question 5

Auto scaling on AZ and Load Balancing across AZ are characteristics of High Availability on AWS?

Answer 5

YES

(Page 117)

Question 6

Elasticity?

Answer 6

Depends on a Scalable System
This means that it (system) has the capacity of “auto-scale” according to the load.

(Page 118)

Question 7

Agility?

Answer 7

The capacity of creating new IT resources is only one click away, for example.

How easy it’s to create new IT resources in few time.

Don’t get confused with SCALABILITY.

(Page 118)

Question 8

Load Balancers?

Answer 8

The server that forwards internet traffic to multiple servers.

It provides a single DNS endpoint and HTTPS; handles failures, and checks the servers’ health where traffic is forwarded.

(Page 119 & 120)

Question 9

In AWS, is AZ available on Load Balancer?

Answer 9

Yes, you can implement an LB along several AZ.

(Page 120)

Question 10

ELB?

Answer 10

Elastic Load Balancer (ELB) that is a Managed Load Balancer.
Warranty Elasticity and Scalability.
Helps to get low costs, reducing the amount of EC2 Instances according to the workload.

(Page 121)

Question 11

What are the three types of ELB?

Answer 11

Application LB
Network LB
Classic/Hybrid LB, which is deprecated

(Page 121)

Question 12

Enum from higher to lower the three different LB according to their performance

Answer 12

1. Network LB (Layer 4), used for High performance
2. Application/HTTP LB (Layer 7)
3. Classic LB, the slowest but work with layers 4 and 7.

(Page 121)

Question 13

Auto Scaling Group (ASG)?

Answer 13

Perform the next actions according to the load:

1. Add out (increase the number of instances) automatically.
2. Add in (decrease the number of instances) automatically.
3. Ensure having the minimum and maximum numbers of EC2 instances
4. Register the EC2 into the LB.
5. Replace unhealthy instances.

(Page 122)

Question 14

What are the Scaling Strategies?

Answer 14

1. Manual Scaling
2. Dynamic Scaling by Simple/Step Scaling
3. Dynamic Scaling by Target Tracking Scale
4. Dynamic Scaling by Scheduling
5. Predictive Scaling

(Page 125 & 126)

Question 15

What is the ASG Manual Scaling Strategy?

Answer 15

Update the size of and ASG manually.

(Page 125)

Question 16

What is the ASG Dynamic Scaling by Simple/Step Scaling?

Answer 16

Adjust the instnances according to some thresholds or CloudWatch, like Processor pass some percent of use, in both ways.

(Page 125)

Question 17

What is the ASG Dynamic Scaling by Target Tracking Scaling?

Answer 17

To set some indicators into a specific range, for example, to keep the CPU processor around 40% of usage

(Page 125)

Question 18

What is the ASG Dynamic Scaling by Scheduled Scaling?

Answer 18

Useful when you know the amount of work around some schedule. So this means that ELB increase or decrease EC2 instances around that schedule.

(Page 125)

Question 19

What is the ASG Predictive Scaling?

Answer 19

Based on Machine learning, adjust the amount of EC2 Instances according to the behavior of the load.

“Predict the Feature”

Useful when the load meets some pattern.

(Page 126)