HA Archictecture

Question 1

What are the three type of elastic load balancer available on aws?

Answer 1

Application load balancers.
Network load balancers.
Classic load balancers.

Question 2

What layer do application load balancers operate on?

Answer 2

Layer 7.

Question 3

What layer do network load balancers operate on?

Answer 3

Layer 4

Question 4

What layers do classic load balancers operate on?

Answer 4

Layer 4 and 7.

Question 5

Network load balancers are used in what use cases?

Answer 5

Where extreme performance is required and for applications that only use tcp to communicate.

Question 6

For classic load balancers, what error code is returned if your application or database fails to respond?

Answer 6

504 - gateway timeout

Question 7

What header can be used to obtain the clients ipv4 address when receiving traffic from an elb?

Answer 7

X-Forwarded-For

Question 8

Sticky sessions for classic load balancers, they allow you to…

Answer 8

Bind a user’s session to a specific ec2 instance.

Question 9

How do sticky sessions differ in application load balancers to classic ones?

Answer 9

The traffic is routed to the target group level instead of the instance level.

Question 10

When you come across a problem where you have a load balancer always having traffic directed to one instance or target group, what should you check first?

Answer 10

Rule out sticky sessions by checking if it is disabled or not.

Question 11

What does cross zone load balancing allow you to achieve?

Answer 11

It allows you to route traffic from one elb from one availability zone to instances in another availability zone.

Question 12

Path patterns allow you to…

Answer 12

Route traffic to certain ec2 instances or micro services based on the url path, eg images calling to another ec2 where application code to executes on a different route of traffic.

Question 13

Auto Scaling consists of three components these are..

Answer 13

Groups - web svr group or app svr group.
Configuration templates - consist of launch templates or launch configurations that specify the spec of the instances to be creating in the auto scaling group.
Scaling options - provides the options of how you want to your auto scaling group to scale, this can be dynamic and rely on conditions or be done on a scheduled basis.

Question 14

What are the five auto scaling options?

Answer 14

Maintain current instance levels at all times.
Scale manually.
Scale based on a schedule.
Scale based on demand.
Scale based on predictive scaling.

Question 15

When you delete your auto scaling group all instances running in that group will…

Answer 15

Terminate if they were created by that auto scaling group.

Question 16

What are the key six rules to remember when considering a HA architecture?

Answer 16

1 always design for failure
2 use multiple az’s and regions where possible
3 know the difference between multi az and read replicas for rds.
4 know the difference between scaling up and scaling out.
5 always assess the cost element and understand it.
6 know the different s3 storage classes.

Question 17

What is the difference between scaling up and scaling out?

Answer 17

Scaling up - increasing the resources at the instance level to handle load.
Scaling out - creating more instances to handle load.