Cisco cloud corner on AWS Elastic Load Balancer (ELB) and Auto Scaling

Prerequisites:

• You must have a valid AWS account

This workshop has been inspired by the aCloudGuru course Deep Dive into Application Load Balancer (ALB)

Creating an Application Load Balancer (ALB). Time: 5 min

An Application Load Balancer makes routing decisions at the application layer (HTTP/HTTPS), supports path-based routing, and can route requests to one or more ports on each EC2 instance or container instance in your VPC.

1. From the AWS console, go to Compute -> EC2
2. Select Load Balancers in the LOAD BALANCING section
3. Select Create Load Balancer
4. Select Application Load Balancer and click Continue
   1. Step 1 - Configure Load Balancer. Select:
      1. Basic Configuration
         • Name: cloudCornerALB
         • Scheme: internet-facing
         • IP address type: ipv4 (I know, it’s bad)
      2. Listeners
         • HTTP: 80
      3. Availability Zones
         • VPC: Default VPC
         • Add all the availability Zones
      4. Click Next:Configure Security Settings
   2. Step 2 - Configure Security Settings
      1. Click Next:Configure Security Groups
   3. Step 3 - Configure Security Groups
      1. Create a new security group
         • Security group name: aCloudCorner-alb
         • Description: acloudCorner-alb
      2. Keep the default rule:
         • Type : Custom TCP Rule
         • Protocol: TCP
         • Port Range: 80
         • Source: 0.0.0.0/0
      3. Click Next:Configure Routing
   4. Step 4 - Configure Routing
      1. Target Group
         • Target group: New target group
         • Name: webServer-tg
         • Protocol: HTTP
         • Port: 80
      2. Health checks
         • Protocol: HTTP
         • Path: /
      3. Advanced health check settings
         • Port: traffic port
         • Healthy threshold: 2
         • Unhealthy threshold: 2
         • Timeout: 5
         • Interval: 6
         • Success codes: 200-299
      4. Click Next:Register Targets
      5. Step 5 - Register Targets
      6. Click Next:Review
      7. Step 6 - Review
      8. Click Create
5. Summary
   • ALB is running
   • ALB listens on port 80 thanks to target group webServer-tg
   • Security Group aCloudCorner-alb allows incoming traffic to ALB on port 80

Creating EC2 with Application Load Balancer (ALB). Time: 10 min

Create an EC2 instance

1. From the AWS console, go to Compute -> EC2
2. Select Instances in the INSTANCES section
3. Select Launch instance, choose the first AMI in the list (Amazon Linux AMI) and click Select
4. Keep the default Instance type (Free tier eligible) and click Next:Configure Instance Details
5. Select:
   • Number of instances: 1
   • Network: default network
   • Subnet: No preference

6. Open the Advanced Details section and copy paste

   #!/bin/bash
   yum update -y
   yum install httpd -y
   service httpd start
   chkconfig httpd on
   echo "Hello Cloud Corners. My Web Server address is: " > /var/www/html/index.html
   curl http://169.254.169.254/latest/meta-data/local-ipv4 >>  /var/www/html/index.html

7. Click Next:Add Storage
8. Click Next:Add Tags and Enter:
   • Key: Name
   • Value: myWebServer
9. Click Next:Configure Security Group
10. Create a new Security Group and Enter:
    • Security Group Name: aCloudCorner-web
    • Description: aCloudCorner-web
11. Click Add Rule and Enter:
    • Type: HTTP
    • Source: sg-xxxxxxxx - aCloudCorner-alb
12. Click Review and Launch
13. Click Launch
14. In the “Select an existing key pair or create a new key pair window”
    • Select an existing key pair or create a new one
    • Acknowledge and click Launch Instance
15. Click View instances. The first EC2 instance is now under creation. When it will be ready, it will have a Public IP assigned
16. Copy/Paste the public IP Address in your browser. Nothing happens because the EC2 instance is not internet accessible. We need to register the EC2 instance with the target group of the load balancer.

Register the EC2 instance with the target group of the load balancer

1. From the AWS console, go to Compute -> EC2
2. Select Target Groups in the LOAD BALANCING section
3. Select webServer-tg target group, go to folder Targets and click Edit
4. Select EC2 instance myWebServer and click Add to registered on port 80
5. Click Save
6. Select Load Balancers in the LOAD BALANCING section
7. Select cloudCornerALB Load Balancer and copy DNS name
8. Paste the ALB DNS name in your browser. You should see the message “Hello Cloud Corners. My Web Server address is: xx.xx.xx.xx”

Cleanup

1. From the AWS console, go to Compute -> EC2
2. Select Instances in the INSTANCES section
3. Select myWebServer instance in the list and click Actions -> Instance State -> Terminate

Autoscaling: Automation, Scalability & Availability. Time: 40 min

Create a Launch Configuration

A launch configuration is a template that an Auto Scaling group uses to launch EC2 instances.

1. From the AWS console, go to Compute -> EC2
2. Select Auto Scaling Groups in the AUTO SCALING section
3. Select Create Auto Scaling Group
4. Select Create launch configuration, choose the first AMI in the list (Amazon Linux AMI) and click Select
5. Keep the default Instance type (Free tier eligible) and click Next:Configure details
6. Select:
   • Name: aCloudCorner-lc

7. Open the Advanced Details section and Enter

   • User Data:

   #!/bin/bash
   yum update -y
   yum install httpd -y
   service httpd start
   chkconfig httpd on
   echo "Hello Cloud Corners. My Web Server address is: " > /var/www/html/index.html
   curl http://169.254.169.254/latest/meta-data/local-ipv4 >>  /var/www/html/index.html

   • IP Address Type: Assign a public IP address to every instance
8. Click Next:Add Storage
9. Click Next:Configure Security Group
10. Select existing Security Group:
    • Name: aCloudCorner-web
11. Click Review
12. Click Create launch configuration
13. In the “Select an existing key pair or create a new key pair window”
    • Select an existing key pair or create a new one
    • Acknowledge and click Create launch configuration

Create an Auto Scaling Group

An Auto Scaling group contains a collection of EC2 instances that share similar characteristics and are treated as a logical grouping for the purposes of instance scaling and management

1. Enter
   • Group Name: aCloudCorner-asg
   • Group size: Start with 2 instances
   • Network: Select the default VPC
   • Subnet: Select all subnets in the list
2. Open the Advanced Details section and Enter:
   • Load Balancing: Receive traffic from one or more load balancers
   • Classic Load Balancers: Let it blank
   • Target groups: webServer-tg
   • Health Check Type: ELB
3. Select Next: Configure scaling policies.
4. Select Use scaling policies to adjust the capacity of this group and Enter
   • Scale between 2 and 3 instances. These will be the minimum and maximum size of your group.
   • Increase Group Size
     • Name: Increase Group Size
     • Execute policy when.
     • Select Add new alarm
       • Unselect Send a notification to
       • Whenever: Average of CPU Utilization Is >= 5 Percent
       • For at least 1 consecutive period(s) of 1 Minute
       • Name of alarm: aCloudCorner-CPUAlarm-UP
       • Select Create Alarm
     • Take the action: Add 1 Instances when 5 <= CPUUtilization < +infinity
       Instances need: 60 seconds to warm up after each step
   • Decrease Group Size
     • Name: Decrease Group Size
     • Execute policy when.
     • Select Add new alarm
       • Unselect Send a notification to
       • Whenever: Average of CPU Utilization Is <= 1 Percent
       • For at least 1 consecutive period(s) of 1 Minute
       • Name of alarm: aCloudCorner-CPUAlarm-DOWN
       • Select Create Alarm
     • Take the action: Remove 1 Instances when 1 >= CPUUtilization > -infinity
5. Select Next: Configure Notifications
6. Select Next: Configure tags and Enter:
   • Key: Name
   • Value: myWebServer-asg
7. Select Review

8. Select Create Auto Scaling group

   The Auto scaling group has now started 2 EC2 instances for you

9. Select Target Groups in the LOAD BALANCING section
10. Select webServer-tg target group, go to folder Targets
11. Refresh your AWS console until the 2 EC2 instances are healthy
12. Got to LOAD BALANCER and click Load Balancers
13. Select cloudCornerALB Load Balancer and copy DNS name
14. Paste the ALB DNS name in your browser. You should see the message “Hello Cloud Corners. My Web Server address is: xx.xx.xx.xx”
15. Refresh the browser multiple times. You should see 2 different IP addresses

Check High Availability

Auto scaling is able to replace a failed EC2 instance

1. From the AWS console, go to Compute -> EC2
2. Select Instances in the INSTANCES section
3. Select one of the running Instances
4. Copy the Public IP
5. Open a terminal
6. Make sure your SSH key is available and is protected against write (chmod 0600)

7. Enter

   ssh ec2-user@<publicIP> -i <keyPair.pem>

   sudo su
   # Stop the HTTP server
   service httpd stop

8. Select Load Balancers in the LOAD BALANCING section
9. Select cloudCornerALB Load Balancer and copy DNS name
10. Paste the ALB DNS name in your browser. You should see the message “Hello Cloud Corners. My Web Server address is: xx.xx.xx.xx”
11. Refresh the browser multiple times. You should see now 1 IP address instead of 2
12. From the AWS console, go to Compute -> EC2
13. Select Target Groups in the LOAD BALANCER section
14. Select the target group webServer-tg in the list

15. Select the Targets tab

    One of the EC2 instance should be “unhealthy”

16. Refresh the AWS console until a new EC2 instance has been created

17. Go back to the DNS browser window
18. Refresh the browser multiple times. You should see now 2 different IP addresses

Automatic Scaling UP based on CloudWatch alarm

We will create a separate EC2 instance to simulate CPU overload on the EC2 instances of the target group

1. From the AWS console, go to Compute -> EC2
2. Select Instances in the INSTANCES section
3. Select Launch instance, choose the first AMI in the list (Amazon Linux AMI) and click Select
4. Keep the default Instance type (Free tier eligible) and click Next:Configure Instance Details
5. Select:
   • Network: default network
   • Subnet: No preference

6. Open the Advanced Details section and copy paste

   #!/bin/bash
   yum update -y
   yum install httpd -y

7. Click Next:Add Storage
8. Click Next:Add Tags and Enter:
   • Key: Name
   • Value: LoadTest
9. Click Next:Configure Security Group
10. Select existing Security Group:
    • Name: aCloudCorner-web
11. Click Review and Launch
12. Click Launch
13. In the “Select an existing key pair or create a new key pair window”
    • Select an existing key pair or create a new one
    • Acknowledge and click Launch Instance
14. Click View instances. The LoadTest EC2 instance is now under creation. When it will be ready, it will have a Public IP assigned. Copy the Public IP address
15. Open a terminal
16. Make sure your SSH key is available and is protected against write (chmod 0600)

17. Enter

    ssh ec2-user@<publicIP> -i <keyPair.pem>

    sudo su
    # We’re using the following command for the load test, to send 500,000 requests to the load balancer using two threads. <UrlLoadBalancer> is replaced by the URL of the load balancer:
    ab -n 500000 -c 2 <UrlLoadBalancer>

18. From the AWS console, go to Management Tools -> CloudWatch
19. Select Alarms
20. Wait 1 or 2 minutes until the alarm aCloudCorner-CPUAlarm-UP is triggered
21. From the AWS console, go to Compute -> EC2
22. Select Target Groups in the LOAD BALANCER section
23. Select the target group webServer-tg in the list

24. Select the Targets tab

    One new EC2 instance should appears with Status “unhealthy”, then “healthy”

25. Refresh the AWS console until a new EC2 instance has been created

26. Go back to the DNS browser window
27. Refresh the browser multiple times. You should see now 3 different IP addresses

Automatic Scaling DOWN based on CloudWatch alarm

1. Go back to the terminal window and enter Ctrl C if the load test has not been ended
2. From the AWS console, go to Management Tools -> CloudWatch
3. Select Alarms
4. Wait 1 or 2 minutes until the alarm aCloudCorner-CPUAlarm-DOWN is triggered
5. From the AWS console, go to Compute -> EC2
6. Select Target Groups in the LOAD BALANCER section
7. Select the target group webServer-tg in the list

8. Select the Targets tab

   One new EC2 instance should appears with Status “unhealthy”

9. Refresh the AWS console until an existing EC2 instance has been removed

10. Go back to the DNS browser window
11. Refresh the browser multiple times. You should see now only 2 different IP addresses

Cleanup

1. From the AWS console, go to Compute -> EC2
2. Select Auto Scaling Groups in the AUTO SCALING section
3. Select the auto scaling group aCloudCorner-asg
4. Select Actions -> Delete
5. Select Launch Configurations
6. Select the launch configuration aCloudCorner-lc
7. Select Actions -> Delete launch configuration
8. Got to INSTANCES and click Instances
9. Select your running EC2 instances in the list and click Actions -> Instance State -> Terminate
10. Got to LOAD BALANCING and click Load Balancers
11. Select the load balancer aCloudCorner-alb
12. Select Actions -> Delete
13. Got to LOAD BALANCING and click Target Groups
14. Select the load balancer webServer-tg
15. Select Actions -> Delete
16. Got to NETWORK & SECURITY and click Security Groups
17. Select aCloudCorner-web and aCloudCorner-alb in the list and click Actions -> Delete Security Groups