Route 53: Health Checks.

 

Amazon Route 53 – Health Checks are a feature used to monitor the health and availability of resources (e.g., web servers, applications, endpoints) and route traffic accordingly based on that health status.

Here’s a breakdown of what Route 53 Health Checks do and how they work:

 Purpose of Health Checks

Route 53 health checks determine if a specified endpoint (like a web server or API) is healthy. If not, Route 53 can:

• Stop routing traffic to it (failover).
• Send alerts using CloudWatch.
• Enable high availability using features like DNS failover.

 What Can Be Monitored

twtech may create health checks for:

• Endpoints (IP addresses or domain names).
• CloudWatch Alarms (to monitor custom metrics).
• Other health checks (calculated health checks aggregate the status of others).

 Types of Health Checks

1. Endpoint Health Checks
• HTTP, HTTPS, or TCP.
• Checks response codes and connection success.
2. Calculated Health Checks
• Combine multiple health checks using logic (e.g., 2 of 3 must be healthy).
3. CloudWatch Alarm Health Checks
• Use a CloudWatch alarm state as the health indicator.

 Configurable Parameters

When creating a health check, twtech would configure:

• IP address or domain to check.
• Port and protocol (HTTP, HTTPS, TCP).
• Path for HTTP/HTTPS checks (e.g., /status).
• Expected response codes.
• Check interval (default 30s; can be 10s for faster checks).
• Failure threshold (number of failed checks before marked unhealthy).
• Optional latency and regional endpoint considerations.

 Health Checkers

Route 53 uses multiple health checker servers located around the globe to reduce false positives and increase accuracy.

 Integration with DNS Routing Policies

Health checks are often used with:

• Failover routing – switch to a backup resource if primary is unhealthy.
• Latency-based routing – only route to healthy endpoints with the lowest latency.
• Geolocation/Geoproximity – combined with health to ensure availability per region.
• Weighted routing – distribute traffic only to healthy endpoints based on weights.

 Notifications

Route 53 can trigger Amazon CloudWatch alarms, which can then send SNS notifications if a health check fails.

 Limitations: What Health Checks Cannot Do

• They do not monitor hosted zones or DNS names—only endpoints.
• They do not automatically restart instances or fix errors (though they can trigger workflows that do).

twtech Use Case

• You're running a web app across us-east-1(Ohio)  and us-west-2(N. California).

 Create:

• Health checks on both endpoints.
• A failover routing policy in Route 53.
• If us-east-1(Ohio) becomes unhealthy, Route 53 routes traffic to us-west-2(N. California).

Project: Hands-on

How twtech creates and use Health check to monitor its resources across the regions.

Go to rout53 console and create health checks for all the instances across the regions.

Create Health checks

Health check on Instance in:  us-east-2

NB:

Follow the same steps to create health checks for the rest of the instances across the regions.

Health check on Instance in:  us-east-1

Health check on Instance in: us-west-1

Exiting and new Health checks created: 

How twtech created a failing health check for it instance by editing the security group of the instance and removing the traffic port. (delete port80).

Go to instance in any of the regions and edit the security inbound rule: 

Edit security inboud rules: delete the traffic port for the application.

From:

To: delete the http port for the application

Save changes:

Port80 that allow traffic to the application in us-east-2 has been removed: to trigger health failure.

Go back to health check in route53 to verify that the instance in us-east-2 is giving an: unhealthy status

From:

To :

Refresh page:  unhealthy

More information can be go from the health checkers:

Select a Health check then, Click on it, to get to: Health checkers

Project: Hands-on

How twtech creates a : calculated Health check for its instances

Project: Hands-on

How twtech creates for its instances a: State of the cloudwateh alarm 

twtech-insights:

If one of the instance display an unhealthy status, that would automatically make the calculated Heath check unhealthy too.