You are deploying over 50 EC2 in separate regions to support your new mobile photo editing app. All have implemented health checks for all Application Load Balancers, as well as CloudWatch alarms for key performance metrics. You have also created Route 53 endpoint health checks for each individual instance. You are growing concerned about the number of SNS notifications you will receive for each of these health checks.
Your ELB health checks can monitor individual instances, but how can you quickly determine whether enough instances in your production environment are unhealthy that it will negatively impact your application’s availability?
Explanation
Implementing a Route 53 Calculated Health Check is the best solution in this case because it aggregates all of the individual health checks. Integrating QuickSight and CloudTrail will create an effective dashboard, but this is not an automated solution because someone has to continuously check the dashboard.
You can also create a CloudWatch Log log alarm, but this would be overly complex compared to the Calculated Health Checks.
For example, let’s say you have deployed 15 EC2 instances, and created 15 separate Route 53 health checks to monitor each one. While you are concerned about each instance’s performance, you essentially need 10 of the 15 instances to be healthy for your system to run properly. Rather than setting up SNS notifications for each individual health check, you can create a calculated health check that collects the results of each health check into a single count between 0-15, and compares that count to a threshold you’ve set. In this case, the threshold could be 10. Now, if 10 instances are healthy, the calculated health check comes back healthy. If less than 10 instances are healthy, the health check comes back unhealthy.For example, let’s say you have deployed 15 EC2 instances, and created 15 separate Route 53 health checks to monitor each one. While you are concerned about each instance’s performance, you essentially need 10 of the 15 instances to be healthy for your system to run properly. Rather than setting up SNS notifications for each individual health check, you can create a calculated health check that collects the results of each health check into a single count between 0-15, and compares that count to a threshold you’ve set. In this case, the threshold could be 10. Now, if 10 instances are healthy, the calculated health check comes back healthy. If less than 10 instances are healthy, the health check comes back unhealthy.
Explanation
Implementing a Route 53 Calculated Health Check is the best solution in this case because it aggregates all of the individual health checks. Integrating QuickSight and CloudTrail will create an effective dashboard, but this is not an automated solution because someone has to continuously check the dashboard.
You can also create a CloudWatch Log log alarm, but this would be overly complex compared to the Calculated Health Checks.
For example, let’s say you have deployed 15 EC2 instances, and created 15 separate Route 53 health checks to monitor each one. While you are concerned about each instance’s performance, you essentially need 10 of the 15 instances to be healthy for your system to run properly. Rather than setting up SNS notifications for each individual health check, you can create a calculated health check that collects the results of each health check into a single count between 0-15, and compares that count to a threshold you’ve set. In this case, the threshold could be 10. Now, if 10 instances are healthy, the calculated health check comes back healthy. If less than 10 instances are healthy, the health check comes back unhealthy.For example, let’s say you have deployed 15 EC2 instances, and created 15 separate Route 53 health checks to monitor each one. While you are concerned about each instance’s performance, you essentially need 10 of the 15 instances to be healthy for your system to run properly. Rather than setting up SNS notifications for each individual health check, you can create a calculated health check that collects the results of each health check into a single count between 0-15, and compares that count to a threshold you’ve set. In this case, the threshold could be 10. Now, if 10 instances are healthy, the calculated health check comes back healthy. If less than 10 instances are healthy, the health check comes back unhealthy.
