Rackspace’s Cloud Load Balancers service includes a health monitoring operation that helps keep your load balancer operating smoothly by routing traffic only to nodes that are functioning properly. You set the service by updating the load balancer configuration settings to add one health monitor per load balancer. After configuration, the monitor periodically checks the health of each node associated with the load balancer, including new nodes that you add. If the health monitor detects an unresponsive node, the system removes it from the load balancer rotation until the health monitor determines that the node is functional.

Configure a health monitor

1. Log in to the Cloud Control Panel.

2. In the top navigation bar, click Select a Product > Rackspace Cloud.

3. Select Networking > Load Balancers.

4. Click the gear icon next to the load balancer name, and select Edit Health Monitoring.

5. In the pop-up dialog box, select a Monitor Type:

   • CONNECT: The Connect health monitor connects to each node on its defined port to ensure that the service is listening correctly. The Connect monitor is the most basic health check and does no post-processing or protocol-specific health checks. Required information for this monitor includes the following details:

     

     • Interval: Minimum number of seconds to wait before executing the health monitor. Must be between 1 and 3600.

     • Timeout: Maximum number of seconds to wait for a connection to be established before timing out. Must be between 1 and 300.

     • Attempts: Number of permissible monitor failures before removing a node from rotation. Must be between 1 and 10.

   • HTTP: The HTTP health monitor is more intelligent than the Connect monitor. It can process an HTTP response to determine the actual condition of a node. Provide the required information for this monitor.

     

     • Interval: Minimum number of seconds to wait before executing the health monitor. Must be between 1 and 3600.

     • Timeout: Maximum number of seconds to wait for a connection to be established before timing out. Must be between 1 and 300.

     • Attempts: Number of permissible monitor failures before removing a node from rotation. Must be between 1 and 10.

     • HTTP Path: The HTTP path for the sample request.

     • Status Regex: A regular expression used to evaluate the HTTP status code returned in the response. For example, you could use the regular expression ^(500|40[1348])$ to look for unsuccessful status codes (500, 401, 403, 404, and 408) returned in the response, or you could use the regular expression ^[2][0][02]$ to look for successful status codes of 200 and 202 in the response.

     • Body Regex: A regular expression used to evaluate the contents of the body of the response. For example, you could use the regular expression ^.*(Unauthorized|Forbidden|Not   Found|Timeout|Server Error).*$ to look for any of those potentially problematic strings in the body of the response, or you could use the regular expression ^success$ to look for the string success.

       Note: The system evaluates only the first 2048 bytes of the response against the body regex you specify, so you should test accordingly.

6. Click Save Monitoring Settings to apply your changes.

7. To debug the HTTP health monitoring, test the body regex against the IP address of the nodes that you disabled. You can use the following cURL command to see what the health monitoring analyzes:

   $ curl -s -r 0-2048 https://YOUR_IP_ADDRESS | head -c 2048 | egrep “YOUR_REGULAR_EXPRESSION”

Disable a health monitor

Use the following steps to disable a health monitor:

1. Click the gear icon next to the load balancer you want to update, then select Edit Health Monitoring.

2. In the health monitor form, click Disable.

3. Click Disable Health Monitor to apply the change.