Module 9: Implementing Elasticity, High Availability, and Monitoring

Question 1

Which statement about Amazon EC2 Auto Scaling is accurate?

• It requires the customer to purchase Reserved Instances.
• It can launch Amazon EC2 instances in multiple Availability Zones
• It can launch Amazon EC2 instances, but customers must terminate instances after they are no longer needed.
• It can only launch new Amazon EC2 instances based on a schedule.

Answer 1

• It can launch Amazon EC2 instances in multiple Availability Zones

Question 2

You detected that the demand on a fleet of Amazon EC2 instances in an Auto Scaling group increases by a set amount each day. Which type of scaling is the most appropriate?

• Scheduled
• Dynamic
• Predictive
• Manually

Answer 2

• Scheduled
  ( Scheduled scaling is most appropriate when you have a well-established pattern of demand. The answer is not “Predictive”. Predictive scaling could be an option, but it is more appropriate when you must discover the pattern for usage demand. It uses artificial intelligence to predict future demand based on historical data.)

Question 3

A fleet of Amazon EC2 instances are launched in an Auto Scaling group behind an Elastic Load Balancing load balancer. The EC2 instances must maintain 50 percent average CPU utilization. Which type of scaling provides the simplest way to achieve this requirement?

• Step scaling
• Simple scaling
• Target tracking scaling
• Manual scaling

Answer 3

• Target tracking scaling
  (Target tracking scaling provides average CPU utilization as a standard target metric. You can specify the target value and let Amazon EC2 Auto Scaling handle the rest)

Question 4

How do you vertically scale an Amazon Relational Database Service (Amazon RDS) database?

• By adding read replicas
• By creating dedicated read and write nodes
• By sharding the database
• By changing the instance class

Answer 4

• By changing the instance class
  ( Changing the instance types with greater capacity is an example of vertical scaling. The answer is not “By adding read replicas…”.
  (Adding read replicas is an example of horizontal scaling.)

Question 5

How do you horizontally scale an Amazon Aurora database?

• By adding Aurora Replica instances
• By increasing the size of the buffer cache configuration
• By creating Amazon CloudWatch alarms
• By changing the instance type

Answer 5

• By adding Aurora Replica instances

Question 6

How does Amazon DynamoDB perform automatic scaling?

• It adds and removes database instances in response to changes in traffic.
• It adds read replicas in response to increased read demand.
• It adjusts the provisioned throughput capacity in response to traffic patterns.
• It changes the instance type in response to changes in processing load.

Answer 6

(The answer is not “ It adds read replicas…”
Amazon DynamoDB does not use read replicas. Relational databases use read replicas.

The answer is not “It adds and removes database instances…”
Amazon DynamoDB does not use instances. Amazon Relational Database Service <Amazon> uses instances.)</Amazon>

Question 7

A fleet of Amazon EC2 instances launch in an Auto Scaling group. The instances run an appication that uses a custom protocol on TCP port 42000. Connections from client systems on the internet must balance across the instances. Which load balancing solution ensures the highest availability?

• Round-robin DNS
• Application Load Balancer
• Network Load Balancer
• Instance-based load balancer

Answer 7

• Network Load Balancer
  ( A Network Load Balancer can handle any TCP, UDP, and TLS traffic. As a feature of Elastic Load Balancing, a Network Load Balancer is highly available.

The answer is not “Application Load Balancer.
As, an Application Load Balancer handles HTTP and HTTPS. It does not work for use case because the application does not use either of these protocols.)

Question 8

Users in location A connect to an application in Region A. Users in location B connect to the same application in Region B. If the application in Region A becomes unhealthy, clients in location A must be redirected to the application in Region B. Which solution can meet this requirement?

• Use an Application Load Balancer with Amazon CloudWatch alarms
• Use geolocation routing with failover records in Amazon Route S3.
• Use latency-based routing in Amazon Route S3 with Amazon CloudWatch alarms.
• Use geoproximity routing and a Network Load Balancer that is attached to both Regions.

Answer 8

• Use geolocation routing with failover records in Amazon Route S3.

( Geolocation routing enables the separation of traffic based on location. A failover record that points to the application in Region B enables failover if the application in Region A becomes unhealthy.)

Question 9

A company must build a highly available website that uses server-side scripts to server dynamic HTML. Which solution provides the highest availability for the least cost and complexity?

• An Auto Scaling group launches Amazon EC2 instances, which are served by an Application Load Balancer. DNS name resolution points to the load balancer.
• Amazon S3 hosts the website. DNS name resolution points to the S3 bucket.
• An Auto Scaling group launches Amazon EC2 instances, which are served by a Network Load Balancer. Amazon Route S3 latency-based routing.
• A second web server is deployed in another Region. Amazon Route S3 uses failover routing for disaster recovery (DR).

Answer 9

• An Auto Scaling group launches Amazon EC2 instances, which are served by an Application Load Balancer. DNS name resolution points to the load balancer.

( The Auto Scaling group can automatically launch instances to handle increased load and terminate instances when they become unhealthy or when the load decreases. The Application Load Balancer is highly available and distributes the load across the instances.)

Question 10

You have created an AWS account for your own personal development and testing. You want your account to stay within the AWS Free Tier and to not generate unexpected costs. Which approach will work and requires the least effort?

• Log in to the AWS Management Console each month and check your billing dashboard.
• Create a service control policy (SCP) to restrict all services that are not included in the AWS Free Tier.
• Create an Amazon CloudWatch alarm to send you an email message when the account billing exceeds $).
• Create an Amzaon CloudWatch metric to monitor account billing and limit it to $0.

Answer 10

(The answer is not “ Create a service control policy to restrict all services that are not included in the AWS Free Tier.
This approach involves extensive manual work in researching which services are not included in the AWS Free Tier, and writing an SCP to restrict them.

The answer is not “ Log in to t he AWS Management Console each month and check your billing dashboard”, because this approach requires continual manual effort and diligence. )