By default, PATROL monitors the availability of all system services except for those where the start-up type is disabled. You can change the monitoring properties of the monitored services or add other services to monitor. The following table shows you how the KM monitors each start-up type by default.
Default service monitoring flags
Startup type | Auto restart | Alarm |
---|---|---|
Automatic | 1 | 1 |
Manual | 0 | 0 |
Disabled | 0 | 0 |
To change the default settings for services, select the Configure Service Monitoring menu command from a Services application instance to perform the following tasks:
By default, the Windows KM monitors all services with startup type as 'automatic' or 'manual'.
Thus, Windows KM monitors a service only if the Monitor pconfig variable for the service is set to '1' and the service is not included in the list of the removedServiceList pconfig variable.
When you select the Configure Service Monitoring > Configure Service menu command, after you select the service you want to configure, you are presented with the monitoring options. The following table provides you with names, descriptions, and default values for these options, and the configuration variable associated with each option.
Service monitoring options
Option | Description | Default (yes/no) | Configuration variable |
---|---|---|---|
Restart service when stopped | If you select this option, PATROL automatically attempts to restart the service when it is stopped (only for services with a startup type of Automatic). To use this option, you must also select the option, Generate a PATROL Alarm/Warn when the service is stopped. | Yes | |
Generate a PATROL Alarm/Warn when the service is stopped | By default, when a service is stopped, PATROL generates an Alarm. However, for a particular service, you can specify a Warning instead. This feature is only for services with a startup type of Automatic. | Yes (Alarm) | |
Enable process monitoring for this service | By default, PATROL monitors only whether services are available. To monitor how much memory and CPU a service executable consumes, you must enable process monitoring for the service. When you enable process monitoring, PATROL monitors the service executable process and displays the monitored process beneath the NT_SERVICE application. | No | |
Use specified command to check status of non responsive service | This feature is available for advanced users who have developed custom executables that can determine the status of a service. If you provide such an executable, the value returned by the executable is assigned to the SvcNotResponding parameter. To ensure that an alarm is generated when the service is not responding, you must set the alarm ranges for the SvcNotResponding parameter to correspond to the appropriate value returned by the executable. For example, if the executable returns the value 1 when the service is not responsive, enable the SvcNotResponding Alarm2 as an Alarm and set the alarm range as 1 to 1. | No |
If the services that you are monitoring are not restarted by PATROL as desired, determine the values of the agent configuration variables that affect whether a service is restarted when it goes down. The following table shows the possible combinations of values for these variables and how each combination causes PATROL to restart (yes) or not restart (no) a monitored service when it goes down.
Configuration variable and service restart: combinations
Service configuration variable | Possible values | |||||||
---|---|---|---|---|---|---|---|---|
DisableServiceRestart (global) | 0 | 0 | 0 | 0 | 1 | 1 | 1 | 1 |
AutoRestart (local) | 0 | 0 | 1 | 1 | 0 | 0 | 1 | 1 |
OverrideGlobalServiceRestart (local) | 0 | 1 | 0 | 1 | 0 | 1 | 0 | 1 |
Service is restarted? (yes/no) | No | No | Yes | Yes | No | No | No | Yes |
For more information about these configuration variables, see Managing configuration variables.
For detailed information about the various configuration tasks that you can perform, see: