Question

Difficulty: Very hardRoute 53 DNS Failover

A SysOps administrator is configuring an active-passive failover strategy for a web application. The primary infrastructure is located in the `eu-west-1` region, consisting of an Application Load Balancer (ALB) and an Auto Scaling group of EC2 instances. The secondary failover target is a static maintenance page hosted in an Amazon S3 bucket configured for website hosting in `us-east-1`.

To monitor the health of the primary application, the administrator creates a custom CloudWatch metric in `eu-west-1` that monitors application error rates. They then create a CloudWatch alarm in `eu-west-1` based on this metric.

Finally, the administrator sets up the Route 53 configuration:
- A Route 53 health check that monitors the state of the CloudWatch alarm.
- A primary failover Alias record pointing to the ALB in `eu-west-1`, associated with the Route 53 health check.
- A secondary failover Alias record pointing to the S3 bucket.

During a simulated failure where the error rate spikes, the CloudWatch alarm transitions to the `ALARM` state in `eu-west-1`, but Route 53 continues to route traffic to the primary ALB in `eu-west-1` instead of failing over to the S3 bucket.

Which of the following is the root cause of this issue, and what is the correct resolution?

  1. A
    The custom CloudWatch metric in `eu-west-1` is configured with standard monitoring (5-minute resolution). Route 53 health checks monitoring CloudWatch alarms require detailed monitoring (1-minute resolution) to transition to an unhealthy status. The administrator must recreate the metric with a 1-minute resolution.
  2. B
    The primary Alias record must not be associated with a custom Route 53 health check. Instead, the primary Alias record must have 'Evaluate Target Health' set to 'Yes' and the health check association removed, which allows Route 53 to automatically evaluate the custom CloudWatch alarm associated with the ALB.
  3. Route 53 health checks can only monitor CloudWatch alarms that are created in the `us-east-1` (N. Virginia) region. The administrator must recreate the CloudWatch alarm in `us-east-1` to monitor the metric in `eu-west-1`, and then update the Route 53 health check to reference the new alarm.Answer
  4. D
    Route 53 DNS failover requires the backend database to be configured in a Multi-AZ deployment in `eu-west-1`. Because the database is currently running as a single-AZ instance, Route 53 cannot initiate the DNS failover mechanism. The administrator must convert the database to Multi-AZ.

Answer

Route 53 health checks can only monitor CloudWatch alarms that are created in the `us-east-1` (N. Virginia) region. The administrator must recreate the CloudWatch alarm in `us-east-1` to monitor the metric in `eu-west-1`, and then update the Route 53 health check to reference the new alarm.
Route 53 health checks that monitor CloudWatch alarms can only retrieve the status of alarms located in the `us-east-1` (N. Virginia) region. Although the custom metric is generated in `eu-west-1`, the CloudWatch alarm itself must be configured in `us-east-1` using cross-region metric support for Route 53 to successfully evaluate its state and trigger DNS failover.

Step-by-Step Solution

1
Analyze the Route 53 health check configuration that monitors the CloudWatch alarm.
Identify that the CloudWatch alarm is created in the `eu-west-1` region.
Route 53 health checks can only retrieve alarm status from the `us-east-1` region.
2
Recreate the CloudWatch alarm in the `us-east-1` region.
The new alarm in `us-east-1` is configured to monitor the custom metric originating in `eu-west-1`.
CloudWatch supports alarms that reference metrics in other regions, allowing the alarm to reside in `us-east-1` while monitoring `eu-west-1` resources.
3
Update the Route 53 health check configuration.
The health check now references the newly created CloudWatch alarm in `us-east-1`.
Route 53 is now able to successfully query the state of the alarm and trigger DNS failover to the secondary record when the alarm enters the `ALARM` state.

Key Concept

Route 53 health checks monitoring CloudWatch alarms must use alarms residing in the us-east-1 region, even if the monitored metric is in another region.
Rate this question