Scaling - Auto Scaling

Question 1

What are the types of auto scaling?

Answer 1

EC2 Auto Scaling, Application auto scaling and AWS Auto Scaling

Question 2

What is Amazon EC2 Auto scaling?

Answer 2

setup scaling groups for EC2; perform health checks to remove unhealthy instances

Question 3

What is Application Auto Scaling?

Answer 3

API used to control scaling for resources other than EC2 like Dynamo, ECS, and EMR; common way to scale other resources

Question 4

What is AWS Auto Scaling?

Answer 4

safely scaling multiple AWS resources; predictive scaling console that can manage a collection EC2, ELB and DDB

Question 5

What is an EC2 Auto Scaling group?

Answer 5

automatically horizontally scales out triggered by an event which adds or removes instances that factors availability, cost and system metrics; add or terminate instances to help ensure correct number of instances to handle your application load

Question 6

What are the 4 EC2 Auto Scaling options?

Answer 6

Maintain - keep specific EC2s running, Manual - max, specific or min EC2s running, Schedule - add or remove based on schedule, Dynamic - based on collected metrics

Question 7

What is a launch configuration?

Answer 7

a template Auto Scaling Group uses to launch instances; specifies VPC and subnets, attached to a ELB, defines a health check grace period, AMI are specified; only one launch configuration for a scaling group at a time; can’t change once created

Question 8

What is an EC2 Auto Scaling Policy?

Answer 8

Dynamic auto scaling option can scale based on the target tracking, simple, and step scaling policies

Question 9

What is Target Tracking Policy?

Answer 9

scale based on predefined metric value such as when CPU reaches 70%, scale up

Question 10

What is Simple Scaling Policy?

Answer 10

Waits until health checks and cool down period expires before add new instances

Question 11

What is Step Scaling Policy?

Answer 11

scaling with more logic and sophistication; require Cloudwatch and high/low thresholds for alarms; better than simple scaling if you have only a single scaling adjustment

Question 12

What is the cool down period?

Answer 12

duration given before scaling up or down; default 300 seconds, used in dynamic scaling, manual scaling, but not schedule scaling; prevent addition of too many instances

Question 13

What are the steps for creating an Autoscaling group?

Answer 13

Create an ALB in EC2
Create a SG for the ALB
Create a Launch Template in EC2
Create a new EC2 Key Pair for launch instances
Create a SG between EC2 and ALB
Name launch template, choose AMI and instance type
Create Auto Scaling Group
Add Launch Template to ASG
Modify Group size and Scaling policies
Test horizontal scaling

Question 14

What is a launch template?

Answer 14

similar to launch configuration, but allows for multiple versions of a template; templates are recommended over launch configurations for newer features; allows for template updates and provision both on-demand and spot instances