How-To Create or Edit a service

You can link a service to a host or host group. A host is only monitored if there is service linked to it!

For creating a new service or editing an existing one, proceed as follows:

Open the Administration Web Interface > Configuration > Setup > Services
Click an existing entry to change it or click “Create new service” to define a new service
Select a command from the list of plugins (for example click on "check_mydevice" if using downloaded XML templates) and click on
Enter the necessary settings and click on to save and apply the changes.

The following describes the setting for a service in detail:

Setting	Description
Service name	Name of the service The name shall not contain special characters or spaces and must be unique. The name must be unambiguous, shall not be assigned twice and is case sensitive.
Hosts	A list of hosts on which this service is defined for. Must contain at least one host or group. Enable the "Check hosts" option if you want to link one single or several single hosts to that service. Enable the "Check host groups" option if you want to link a host group to that service. Via double-click in the "Non-members" selection box, add a host or host group to the "Members" selection box or click on it and use the following button. Via double-click in the "Members" selection box, remove a host or host group if you do not want to link it to the service or click on it and use the following button.
Command	Individual settings for that service. The available settings depend on the selected plug-in. Click on to have further information about the selected plug-in.
Normal check frequency (minutes)	An interval in minutes at which a service with OK status shall be checked. If performance data are generated for the service, all collected performance data would be destroyed if the value in the "Normal check frequency" input box was changed. Therefore, the input box is disabled if "Yes" was selected from the "Process performance data" selection box (see below).
Check attempts	Specify how often a service shall be checked in the event of a fault before its status is changed from WARNING to CRITICAL or before notification will be sent out.
Retry check frequency (minutes)	Frequency in minutes for retry check attempts (see above). It define that in the event of a fault, the time interval for a recheck shall be shortened. For example if you check the service every 5 minutes (see "Normal check frequency"), in the event of a fault, the check is not executed every 5 minutes but at the entered time interval (3 minutes, for example).
Timeperiod to perform checks	The time period in which checking is enabled.
Notification frequency (minutes)	Specify at what intervals (in minutes) notifications shall be sent. Enter "0" to only receive one notification in the event of a fault. As long as the status does not change, no further notifications are sent. Only after a status change is another notification sent.
Notification timeperiod	Link alarms of the host status to one of the configured time periods.
Send notifications when Service is in state	Options to determine, when or rather in which fault service state notifications should be sent out.
First notification delay (minutes)	Defines the number of minutes to wait before sending out the first problem notification when this service enters a non-OK state. Value 0 means sending out notifications immediately.
Contactgroups receiving notifications	A contact group to receive a notification in the event of a fault or state change.
Enable flap-detection	Flap detection activation/deactivation. For more information, please view How-To Detect hosts or services Flap.
Lower flapping-threshold	Use to specify the low state change threshold used in flap detection for this host. If you set this directive to a value of 0, the program-wide value specified by the low_host_flap_threshold directive will be used.
Upper flapping-threshold	Use to specify the high state change threshold used in flap detection for this host. If you set this directive to a value of 0, the program-wide value specified by the high_host_flap_threshold directive will be used.
Event handler	Specify, if event handler should be called if a state change occurs. Event handler will send event data to appropriate host. This assumes, that a Sonarplex agent is up and running on destination host. Please make sure that property 'Agent Password' is set for appropriate host (see How-To Create or Edit a host). Event handler has no knowledge about configured check command, therefore password for sending event will be taken from host properties. This applies for check execution too, if check command argument 'Password' is empty, which is the best practice anyway, if appropriate service is configured for more than one host.
Event when service is in state	Specify the fault status in case of which the event handler shall take action.
Global Event Handler	The IP address of a host that provides a central event handler.
Process performance data	Select "Yes" if the performance data of the service shall be saved into database for later graphical evaluation. Select "No" if the performance data of the service shall not be saved or if you want to change the value in the "Normal check frequency" input box (see above).
Performance data type	Specify the type of performance data.
Performance graph type	Specify the type of performance graph.
Notes (optional)	Container for additional service configuration data.
Execute active checks	Specifies if checks should be executed by Sonarplex.
Execute passive checks	Specifies if passive check results should be accepted for this service.
Freshness checking	Specifies if in case of passive checks their fressness should be observed. This means that if freshness threshold is exceeded the specified Command will be executed.
Service is volatile	Use to denote whether the service is "volatile". Services are normally not volatile. Volatile services are resetting their internal state back to "OK" every time they get checked. This means that notifications and events are generated for every non OK result state, even if their previous one was the same. For example when receiving SNMP traps, use this option in combination with "Maximum check commands: 1". This will force a notification for each incoming SNMP Trap, even when they are recognized in short intervals (e.g. each second).
Freshness threshold in seconds	Specifies the interval in seconds a passive service will wait for incoming result before its Command will be executed actively.
My Properties (optional)	Additional properties with the template name, the alias service name and so on.