Chapter 21. Service Monitoring

The Symbiosis system is comprised of several distinct components, which we’ve documented throughout the course of this reference:

Each of these services runs in an independent fashion, and it is possible under certain circumstances that these services might fail, or stop themselves.

To handle the case of services failing to execute normally we’ve included an automated service checker as part of the Symbiosis system. The service checker will check upon the health of your system, by default once every two minutes, and it will automatically restart any services which have failed.

The ‘symbiosis-monit` command is responsible for testing each of the available services, and restarting the failed ones. By default it is executed every two minutes, such that it may respond quickly to failures. It will also stop services that are not required. For example if the machine is not configured to scan any domains’ mail, then SpamAssassin will be stopped.

At any time you wish to check upon the health of your system you may launch it manually, when connected to your server via SSH.

admin@example:~$ sudo symbiosis-monit
=  service test report ========================

 * Host: kvm4.vm.bytemark.co.uk
 * Tests started at: Fri, 11 Jun 2010 15:55:45 +0100

 * apache2: PASS
 * clamav-daemon: PASS
 * clamav-freshclam: PASS
 * cron: PASS
 * dovecot: PASS
 * exim4: PASS
 * mysqld: PASS
 * pure-authd: PASS
 * pure-ftpd: PASS
 * spamassassin: PASS
 * sshd: PASS

 * 11/11 tests passed on first attempt.
 * Tests finished at: Fri, 11 Jun 2010 15:55:46 +0100

= End of service test report ==============================================

In this case all services are working correctly, so "PASS" was reported instead of the failing "FAIL". The possible output status are:

FAIL
The service failed.
PASS
The service appears to be running correctly.