Monitoring health checks using CloudWatch
Route 53 health checks integrate with CloudWatch metrics so that you can do the following:
-
Verify that a health check is properly configured.
-
Review the status of a health check over a specified period of time.
-
Configure CloudWatch to send an Amazon SNS alert when the status of a health check is unhealthy. Note that several minutes might elapse between the time that a health check fails and the time that you receive the associated SNS notification.
For more information, see How Amazon Route 53 determines whether a health check is healthy.
Topics
View the status of your health check
Note
We're updating the health checks console for Route 53. During the transition period, you can continue to use the old console.
Choose the tab for the console you are using.
View health check alarms
Note
We're updating the health checks console for Route 53. During the transition period, you can continue to use the old console.
Choose the tab for the console you are using.
View health check metrics on the CloudWatch console
To view Route 53 metrics on the CloudWatch console
-
Sign in to the AWS Management Console and open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/
. -
Change the current region to US East (N. Virginia). Route 53 metrics are not available if you select any other region as the current region.
In the navigation pane, choose Metrics.
On the All metrics tab, choose Route 53.
Choose Health Check Metrics.
You can also set up SNS notification on the CloudWatch console. For more information, see Create recommended alarms in the CloudWatch User Guide.
Create an alarm with an SNS notification
Note
The following procedure applies to the old console only. The new console directs you to the CloudWatch console to create alarms. For more information, see Find and create recommended alarms in the CloudWatch User guide.
To receive an Amazon SNS notification when a health check status is unhealthy (old console)
In the navigation pane of the Route 53 console, choose Health Checks.
Choose the row for the applicable health check.
-
In the bottom pane, choose the Alarms tab.
The table lists the alarms that you've already created for this health check.
Choose Create Alarm.
-
Specify the following values:
- Alarm name
Enter the name that you want Route 53 to display in the Name column on the Alarms tab.
- Alarm description
(Optional) Enter a description for the alarm. This value appears in the CloudWatch console.
- Send notification
Choose whether you want Route 53 to send you notification if the status of this health check triggers an alarm.
- Notification target (Only when "Send notification" is "Yes")
If you want CloudWatch to send notification to an existing SNS topic, choose the topic from the list.
If you want CloudWatch to send notification but not to an existing SNS topic, do one of the following:
If you want CloudWatch to send email notification – Choose New SNS topic and continue with this procedure.
If you want CloudWatch to send notification by another method – Open a new browser tab, go to the Amazon SNS console, and create the new topic. Then return to the Route 53 console, choose the name of the new topic from the Notification target list, and continue with this procedure.
- Topic name (Only when you choose to create a new Amazon SNS topic)
Enter a name for the new Amazon SNS topic.
- Recipient email addresses (Only when you choose to create a new Amazon SNS topic)
Enter the email address that you want Route 53 to send an SNS notification to when a health check triggers an alarm.
- Alarm target
Choose the value that you want Route 53 to evaluate for this health check:
Health check status – Route 53 health checkers report that the health check is healthy or unhealthy
Health checkers that report the endpoint healthy (%) (health checks that monitor an endpoint only) – The percentage of Route 53 health checkers that report that the status of the health check is healthy
Number of healthy child health checks (calculated health checks only) – The number of child health checks in a calculated health check that report that the status of the health check is healthy
TCP connection time (HTTP and TCP health checks only) – The time in milliseconds that it took Route 53 health checkers to establish a TCP connection with the endpoint
Time to complete SSL handshake (HTTPS health checks only) – The time in milliseconds that it took Route 53 health checkers to complete the SSL/TLS handshake
Time to first byte (HTTP and HTTPS health checks only) – The time in milliseconds that it took Route 53 health checkers to receive the first byte of the response to an HTTP or HTTPS request
- Alarm target
For the alarm targets that are based on latency (TCP connection time, Time to complete SSL handshake, Time to first byte), choose whether you want CloudWatch to calculate latency for Route 53 health checkers in a specific region or for all regions (Global).
Note that if you choose a region, Route 53 measures latency only twice per minute, and the number of samples will be smaller than if you choose all regions. As a result, outlying values are more likely. To prevent spurious alarm notifications, we recommend that you specify a larger number of consecutive periods that the health check must fail before CloudWatch sends you a notification.
- Fulfill condition
-
Use the following settings to determine when CloudWatch should trigger an alarm.
Alarm target Recommended condition Description Health check status
Minimum < 1
Route 53 health checkers report when the endpoint is unhealthy.
Health checkers that report the endpoint healthy (%)
Average < desired percentage
Health checks that monitor an endpoint only – Route 53 considers the status of a health check to be unhealthy when less than 18% of health checkers report that the status is healthy. Don't choose Sample Count for this metric because the range of sample counts can change as Route 53 adds more health checking regions. Average will always accurately represent the percentage of checkers that are reporting the status of a health check.
Number of healthy child health checks
Minimum < desired number of healthy child health checks
The Minimum statistic returns the most conservative value and represents the worst-case scenario.
TCP connection time
Average > desired time in milliseconds
Average is a more consistent value than other statistics.
Time to complete SSL handshake
Average > desired time in milliseconds
Average is a more consistent value than other statistics.
Time to first byte
Average > desired time in milliseconds
Average is a more consistent value than other statistics.
- For at least
x
consecutive periods ofy
minutes/hours/day -
Specify how many consecutive time periods that the specified value must meet the criteria before Route 53 sends notification. Then specify the length of the time period.
-
When you choose Create, Amazon SNS sends you an email with information about the new SNS topic.
-
In the email, choose Confirm subscription. You must confirm your subscription to begin receiving CloudWatch notifications.