Load Balancing

Question 1

What is an ELB?

Answer 1

Elastic Load Balancer, managed by Amazon (AWS takes care of upgrades etc)

Question 2

What are the four types of managed load balancer?

Answer 2

Classic load balancer (CLB) - deprecated. Application load balancer (ALB). Network load balancer (NLB) and Gateway load balancer (GLB).

Question 3

What protocols does the ALB (application load balancer) use?

Answer 3

HTTP, HTTPS & WebSocket

Question 4

What protocols does the NLB (network load balancer) use?

Answer 4

TCP, TLS, UDP

Question 5

What protocols does the GLB (gateway load balancer) use?

Answer 5

IP (operates at network layer 3)

Question 6

Can a ALB support multiple applications?

Answer 6

Yes

Question 7

What can a ALB target group contain?

Answer 7

EC2 instances (can be managed by an auto-scaling group), ECS tasks, Lambda functions, private IP addresses

Question 8

Can the network load balancer have more than one static IP per AZ?

Answer 8

No

Question 9

What protocols does the network load balancer target group health check support?

Answer 9

TCP, HTTP, HTTPS

Question 10

Describe the main use-case for a gateway load balancer?

Answer 10

Network traffic enters the GLB and using a target group is balanced across appliances to monitor traffic (firewall, intrusion detection for example). Traffic exits the GLB and continues to its destination, or the traffic can be dropped.

Question 11

What is SNI?

Answer 11

Server Name Indication - allows using multiple SSL certificates for different ELB target groups / websites

Question 12

What is an ASG?

Answer 12

Auto Scaling Group

Question 13

What is a launch template (for ASG)?

Answer 13

Describes how an EC2 instance should be created by an ASG. Contains attributes like the AMI, instance type, ESB volumes, security groups, IAM role, VPC/SubNet, Load Balancer

Question 14

You have an Auto Scaling Group fronted by an Application Load Balancer. You have configured the ASG to use ALB Health Checks, then one EC2 instance has just been reported unhealthy. What will happen to the EC2 instance?

Answer 14

The ASG will terminate the instance, a new instance will be created in its place.

Question 15

An application is deployed with an Application Load Balancer and an Auto Scaling Group. Currently, you manually scale the ASG and you would like to define a Scaling Policy that will ensure the average number of connections to your EC2 instances is around 1000. Which Scaling Policy should you use?

Answer 15

Target Tracking Policy

Question 16

Describe the target tracking auto scaling policy (ASG)

Answer 16

Define a target value of a metric for example CPU to stay around 40% - cloud watch alarms are automatically created to achieve this.

Question 17

Describe the simple step scaling auto scaling policy (ASG)

Answer 17

Uses cloud watch alarms to scale out or in by a predefined quantity. For example when CPU > 80% then add 2 units.

Question 18

Describe the scheduled scaling auto scaling policy (ASG)

Answer 18

Anticipate scaling based on known usage patterns, for example increase the min capacity on Friday afternoons.

Question 19

Describe the predictive scaling auto scaling policy (ASG)

Answer 19

Continuously forecast load and schedule scaling ahead

Question 20

What are some good examples of metrics to perform autoscaling on (ASG)

Answer 20

CPU utilisation, request count per target, average network in/out, any custom metric pushed to cloud watch

Question 21

What is the auto scaling group cooldown feature?

Answer 21

After an autoscale event the ASG will enter a cooldown phase where it waits for metrics to stabilise before reacting to more autoscale events.