Adding a Health Monitor

Once a load balancer is created with at least one listener and member, Hostwinds clients can then create a health monitor. This article is intended for Hostwinds clients who would like to add a Health Monitor from the Cloud Control.

The health monitor is used to be aware of the health status of all members, so as to properly and reliably route requests to active nodes for processing.

add health monitor

To add a Health Monitor, select the menu option on your listener, to Add Health Monitor.

add health monitor

The above is the pop-up that will display, showing the configuration options for the health monitor being created.

 

Health Monitor Options

Delay

The delay, is the time in milliseconds between sending two consecutive requests to pool members, the default is set to 5.

Up Retries

The number of allowed connection failures before the monitor concludes that the node is inactive, the default setting is 3.

Timeout

The maximum time in milliseconds that a monitor waits for a response from a node before it claims the node is unreachable, the default is set to 5.

Down Retries

The number of retries of a failed member, before the member is marked as “error” status.

Monitor Type

HTTP

The application-side component of an HTTP health check will probe the “/” path on the application server, this however is not appropriate for all solutions, as the path “ / ” may be a cached page. As such, this monitor type has additional fields available.

URL Path

Path part of the URL that should be retrieved from the back-end server, the default is set to “/”

HTTP Method

Select from a range of options to retrieve the url path, default is set to “GET”.

Options available:

  • CONNECT
  • DELETE
  • GET
  • HEAD
  • OPTIONS
  • PATCH
  • POST
  • PUT
  • TRACE

HTTP Response Codes

List of HTTP status codes that indicate an OK health check, the default is set to “200”

Options available:

  • 200
  • 201
  • 202
  • 203
  • 204

HTTPS

Operates exactly as the HTTP monitor type, but with ssl back-end servers. This can cause problems if the servers are performing client certificate validation, in this case using TLS-HELLO type monitoring is an alternative.

TCP

Health monitor opens a TCP connection to the back-end server’s protocol port that you have set. The custom TCP application should be written to respond OK to the load balancer connecting, opening a TCP connection and closing it again after the TCP handshake, and not sending any data.

PING

Health monitor sends a period ICMP PING request to the back-end servers, the server will need to be configured to respond/allow PING in order for this check to pass.

TLS_HELLO

Health monitors ensure the back-end server responds to the SSLv3 client hello messages, it will not check any other metrics or status codes.

Viewing Health Monitors

view health monitors

To view your health monitor, expand the menu option under your Load Balancer’s listener.


Share:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.