Monitoring¶
Monitoring is a module of the Smartswitch system that monitors (checks for functionality) various services on the local host or on remote servers, as well as the network availability of remote servers.
This functionality is used to check the functionality of services on remote and local systems.
When a problem is found, the system sends a warning letter to the configured e-mail address.
The functionality is provided by the daemon application /usr/local/sbin/smartswitch/monitord.
This application, once launched, periodically polls the Monitored Services configured on Hosts.
Observable services contains a bash script that is used to check the functionality of a given service.
This script usually calls external programs that check the service itself.
When the script is called, it is passed the hostname from Hosts.
The system, when installed by default, has several pre-configured Observable services:
- ping - checking the reachability of the host via ping.
A problem is reported if at least 1 ICMP packet was not responded to. - h323 - check the functionality of the H.323 service
- snmp-disk - checking limits on occupied space on the hard drive
- ping-100 - checking the reachability of the host via ping.
A problem is reported if 100% of ICMP packets were not responded to. - snmp-availablility - check the functionality of SNMP on a remote server
- snmp-openfiles - checking the limit on open files on a remote server
After calling the script, the daemon application checks the string that the script returns.
If the line is empty, it is assumed that no error occurred.
If the line is not empty, an error is considered to have occurred, and this line is:
- added to the error database
- added to the letter sent to the administrator.
To notify about an error, Notifiers are used, configured on each Host.
Notifications by e-mail are sent not every time an error is detected, but at a certain period.
The letter to the digest collects all the errors that occurred during the last period.
Otherwise, the administrator will be inundated with spam.
To control the period for sending notifications, notification period (sec) and notification aging timeout (sec) are used.
In addition to sending notifications by e-mail, all errors can be viewed through the Notifications report.