Health check options are set in the Real Servers section in the Virtual Service modify screen in the LoadMaster Web User Interface (WUI) (Virtual Services > View/Modify Services > Modify).

This section lists the Real Servers that are assigned to the Virtual Service. The properties of the Real Servers are summarized and there is also the opportunity to add or delete a Real Server, or modify the properties of a Real Server.

Real Server Check Method

This provides a list of health checks for well-known services, as well as lower level checks for TCP/UDP or ICMP. With the service health checks, the Real Servers are checked for the availability of the selected service. With TCP/UDP, the check is simply a connect attempt.

When one of the HTTP/HTTPS, Generic or STARTTLS protocols Service Types are selected, the following health check options are available:

  • ICMP Ping
  • HTTP Protocol
  • HTTPS Protocol
  • TCP Connection Only
  • Mail (SMTP) Protocol
  • Network News (NNTP) Protocol
  • File Transfer (FTP) Protocol
  • Telnet Protocol
  • Mailbox (POP3) Protocol
  • Mailbox (IMAP) Protocol
  • Binary Data
  • LDAP
  • None

When the Remote Terminal Service Type is selected the following health check options are available:

  • ICMP Ping
  • TCP Connection Only
  • Remote Terminal Protocol
  • None
Note: For a UDP Virtual Service, only the ICMP Ping and Name Service (DNS) Protocol options are available for use.

Check Parameters

In LoadMaster firmware version 7.2.52, the check Interval, Timeout, and Retry Count settings can be configured on each Virtual Service or SubVS. Previously, these were just global settings. You can configure the global settings in Rules & Checking > Check Parameters. The global settings are used by default for all Virtual Services.

Interval (seconds): This field specify the number of seconds that will pass between consecutive health check attempts and defines the length of the ‘health check cycle’. To override the global interval, you can select the any other value from the drop-down list. The global option cannot be selected if either Timeout or Retry Count parameters have been set to a value other than the global setting.

Timeout (seconds): This is the allowed maximum wait time for a connection to be established and for the receipt of a reply to a health check. If no reply is received before the timeout expires, then the connection is re-attempted as long as the Retry Count has not been reached. If any HTTP response is received in response to the health check, then no retries are attempted in the current health check cycle. To override the global timeout value, you can select the any other value from the drop-down list.

Retry Count: This specifies the number of times the health check will be re-attempted if the timeout above is reached before receiving a response. To override the global retry count value, you can select the any other value from the drop-down list. The retry count does not apply if any response is received in response to the health check. If, for example, either a 200 or a 404 response is received, then no retries are attempted in the current cycle.

To configure these settings for a specific Virtual Service, expand the Real Servers section of the Virtual Service or SubVS modify screen. A Real Server Check Method must be selected to see the relevant fields. You can configure these settings to either use the global value, set a specific value within the provided range, or reset to the default value.

If you configure these settings for a parent Virtual Service and then create a SubVS within that Virtual Service, the check values are reset to use the global values.

Checked Port

The port to be health checked. If no port is specified, the port of the Real Server is used.

Note: Extra Ports specified in the Virtual Service are not used for Real Server health checking.

Enhanced Options

Enabling the Enhanced Options check box provides additional health check options:

  • Minimum number of RS required for VS to be considered up
  • Critical
  • Healthcheck On

If the Enhanced Options check box is disabled (the default), the Virtual Service is considered available if at least one Real Server is available. If the Enhanced Options check box is enabled, you can specify the minimum number of Real Servers that must be available to consider the Virtual Service to be available.

Minimum number of RS required for VS to be considered up

Note: This option will only appear if the Enhanced Options check box is enabled and if there is more than one Real Server.

Select the minimum number of Real Servers required to be available for the Virtual Service to be considered up.

If less than the minimum number of Real Servers is available, a critical log is generated and the Virtual Service is marked as down. If some Real Servers are down but it has not reached the minimum amount specified, a warning is logged. If the email options are configured, an email is sent to the relevant recipients. For further information on the email options, refer to the Web User Interface (WUI), Configuration Guide.

Note: Note that the Critical option has a higher priority than the Minimum number of RS required for VS to be considered up option. The system marks a Virtual Service as down whenever a Real Server that is marked as Critical (that is, the Critical check box is enabled) becomes unavailable – even if Enhanced Options are enabled and there are more than the specified minimum number of Real Servers still available.

In all cases, if the Virtual Service is considered to be down and the Virtual Service has a sorry server or an error message configured, these are used. The sorry server is never health checked, it is assumed to be up.

If the minimum number is set to the total number of Real Servers and one of the Real Servers is deleted, the minimum will automatically reduce by one.

A SubVS is said to be available and can be routed to if and only if the number of available Real Servers within that SubVS is greater than or equal to the limit.

If the number of available Real Servers is below this limit, the SubVS will not accept traffic as it is marked as down for failing the health check and this is logged appropriately.

Critical

Note: This option will only appear if the Enhanced Options check box is enabled.

In the Real Servers section of the Virtual Service modify screen, there is a Critical check box for each of the Real Servers. Enabling this option indicates that the Real Server is required for the Virtual Service to be considered available. The Virtual Service is marked as down if the Real Server has failed or is disabled. Therefore, in the example above the parent Virtual Service is down because there is a critical SubVS down. The critical SubVS is down because a critical Real Server is down.

If a Real Server on a SubVS is marked as critical – the SubVS is marked as down if that Real Server is down. However, the parent Virtual Service will not be marked down unless that SubVS is marked as critical.

Note: The Critical option has a higher priority than the Minimum number of RS required for VS to be considered up option. For example, if three Real Servers are added and if the minimum is set to two and only one Real Server is down but that Real Server is set to critical – the Virtual Service is marked as down.

In all cases, if the Virtual Service is considered to be down and the Virtual Service has a sorry server or an error message configured, these are used.

Note: It is not possible to perform complex, chained health checking such as AND or OR statements. For example – if there are five Real Servers, it is not possible to say; IF Real Server 2 is up OR Real Server 4 is up, mark the Virtual Service as up.
Note: You can mark both Real Server 2 and Real Server 4 as critical, but if either of them go down, the parent Virtual Service is marked as down.

Healthcheck On

Note: This option will only appear if the Enhanced Options check box is enabled.
Note: This option is only available on LoadMasters with firmware version 7.1.35 and above.

In the Real Servers section of the Virtual Service modify screen, there is a Healthcheck On drop-down list for each of the Real Servers. This allows you to specify what Real Server the health check is based on. This can either be set to Self in order to perform the health check based on that particular Real Server status, or another Real Server can be selected. For example – if Real Server 1 is down, any Real Servers which have their health check based on Real Server 1 will also be marked as down, regardless of their actual Real Server status.

In the text below, the term “follow” is used to indicate a health check relationship between two Real Servers. For example – if there are two Real Servers called RS1 and RS2. RS1 is said to “follow” RS2 if RS2 is selected in the Healthcheck On drop-down list for the RS1 health check.

Some points to be aware of are listed below:

  • A Real Server can only follow a Real Server and not a SubVS.
  • A Real Server can follow a Real Server that is also following a third Real Server. The status of the first two Real Servers will reflect the status of the third Real Server.
  • Chains of Real Servers are allowed – but loops are not allowed and cannot be created.
  • If a Real Server is deleted (either singly or as part of a Virtual Service), all Real Servers that are following the Real Server are reset to normal behavior (that is, they will start using the Virtual Service health check options).
  • If all Real Servers in a Virtual Service are following Real Servers on a different Virtual Service, the health check parameters for the Virtual Service are not shown on the WUI (because the settings do not affect any Real Servers).
  • Disabling the Enhanced Options check box will disable all Real Server health check following for that Virtual Service and regular checking on the individual Real Servers will resume. If Session Management is enabled, an entry is recorded in the audit log any time the Enhanced Options check box is enabled or disabled.