AWS Route53 - get uptime percentage - amazon-web-services

Question

I need to find a way to report on website uptime (as a percentage) based on AWS Route53 monitoring. This reporting is generally done quarterly.
My initial thought was to have CloudWatch send ALARM and OK states via SNS to SQS, and then process this queue into a database for later reporting. As far as I can tell, however, CloudWatch will only send emails even though an SQS queue is subscribed to the topic.
Any suggestions of how I might achieve this?

Answer 1

Amazon Route 53 can be configured to send healthcheck data to Amazon SQS.
It worked for me -- here's the steps I took:
Create an Amazon SNS notification topic in us-east-1 (where Route 53 performs its health checks)
Create an Amazon SQS queue in us-east-1 (same region as the notification topic)
Subscribe the Amazon SQS queue to the Amazon SNS topic (via the Queue Actions menu option)
Create an Amazon Route 53 Health Check. Set Create Alarm to Yes. Configure it to Send notification to Existing SNS topic and choose the topic created above.
An Amazon CloudWatch alarm will be automatically created by Amazon Route 53.
This will result in health notifications arriving in the SQS queue. However, it will only send an ALARM notification -- there is no notification when it becomes healthy again. To receive a "now healthy" notification, edit the CloudWatch alarm and add a new Notification that triggers when "State is OK".
Here is an example of a failure notification retrieved from the SQS queue:
{
"Type" : "Notification",
"MessageId" : "4768e8e4-0026-51c7-aa6e-a696bf02f808",
"TopicArn" : "arn:aws:sns:us-east-1:123456789012:r53-east",
"Subject" : "ALARM: \"awsroute53--4c2f-9816-a42c50ec8671-High-HealthCheckStatus\" in US - N. Virginia",
"Message" : "{\"AlarmName\":\"awsroute53-4c2f-9816-a42c50ec8671-High-HealthCheckStatus\",\"AlarmDescription\":null,\"AWSAccountId\":\"743112987576\",\"NewStateValue\":\"ALARM\",\"NewStateReason\":\"Threshold Crossed: 1 datapoint (0.0) was less than the threshold (1.0).\",\"StateChangeTime\":\"2015-09-16T00:50:44.591+0000\",\"Region\":\"US - N. Virginia\",\"OldStateValue\":\"OK\",\"Trigger\":{\"MetricName\":\"HealthCheckStatus\",\"Namespace\":\"AWS/Route53\",\"Statistic\":\"MINIMUM\",\"Unit\":null,\"Dimensions\":[{\"name\":\"HealthCheckId\",\"value\":\"4c2f-9816-a42c50ec8671\"}],\"Period\":60,\"EvaluationPeriods\":1,\"ComparisonOperator\":\"LessThanThreshold\",\"Threshold\":1.0}}",
"Timestamp" : "2015-09-16T00:50:44.656Z",
"SignatureVersion" : "1",
"Signature" : "KvCHsBh95q...cw8A==",
"SigningCertURL" : "https://sns.us-east-1.amazonaws.com/SimpleNotificationService-90147a5624348ee.pem",
"UnsubscribeURL" : "https://sns.us-east-1.amazonaws.com/?Action=Unsubscribe&SubscriptionArn=arn:aws:sns:us-east-1:123456789012:r53-east:4b5d-8318-57bd58f0b3a4"
}

Answer 2

One option is to offload Route53 health check statistics to Axibase Time Series Database and enable scheduled reports as discussed in the uptime reports article.
Implementation notes:
Need to create a read-only account to query CloudWatch statistics and Route53 Health Check metadata.
Offload task has a latency of 5-15 minutes (configurable).
Offload task ensures no data gaps in copied CW statistics, i.e. when collection temporarily stops for whatever reason.
Available reports:
Base report: average uptime over the period (day, week, month, quarter).
Enhanced uptime report with additional check configuration details.
Filter results by protocol or custom tag.
Aggregate uptime by custom tag such as GEO, environment.
Filter results by hours of the day or working days.
Aggregate uptime by day of week.
Downtime incident count.
Longest downtime incidents.
The reports can be generated interactively, via a web-based console, delivered via email, or displayed on portals.
Disclaimer: I work for Axibase.