Elastic Load Balancing and Auto Scaling Groups

Question 1

What is another term for horizontal scalability?

Answer 1

Elasticity

Question 2

What is vertical scalability?

Answer 2

Increasing the size of the instance (e.g. more RAM, CPU, etc.)

Question 3

What is the mechanism for horizontal scalability?

Answer 3

Add more instances

Question 4

What does high availability mean?

Answer 4

Running your applications in more than 1 AZ. This builds fault tolerance as if one AZ goes down, you can switch to the failover - thus increasing availability of your resources.

Question 5

What is agility in reference to the tools that the cloud provides?

Answer 5

The idea that they are only one click away which means that you reduce the time to make those resources available to developers from weeks to minutes.

Question 6

What is a load balancer?

Answer 6

Servers that forward traffic to multiple servers downstream, making sure that no one server will take on all of the load

Question 7

Why are load balancers useful?

Answer 7

• Allows us to balance the backend more effectively
• Handles failures of instances effectively
• High availability across zones

Question 8

What are the 3 types of load balancer and what layers do they operate at?

Answer 8

• Application load balancers (layer 7)
• Network load balancers (layer 4)
• Gateway load balancers (layer 3)

Question 9

What is an auto scaling group used for?

Answer 9

To automatically create servers on the backend to adjust for demand. This increases scalability.

Question 10

What are the 4 different dynamic scaling types for an auto scaling group?

Answer 10

• Simple/step scaling (like a decision rule, e.g. if CPU usage > 70%, add 2 units)
• Target tracking scaling (e.g. I want the average CPU usage to be at 40% < x < 50%)
• Schedule scaling (anticipate scaling based on seasonal patterns)
• Predictive scaling (using ML to predict demand ahead of the kerb)