This section explains how to add, configure and utilize Time Periods within Opsview Monitor.
This document explains the concept of Time Periods, what a Time Period is, how to configure a Time Period, what Opsview Monitor items use Time Periods, and how to change these values. Time Periods in the context of notifications are not covered within this document.
After reading the User Guide, users should be able to create their own Time Periods and fine-tune various Opsview Monitor items to use different Time Periods as required.
Time Periods are used within Opsview Monitor for one of two purposes; determining when a Host or Service Check is being actively monitored, and determining when notifications should be sent.
For example, if a Host only needs to be monitored during office hours, then an administrator can create a Time Period called 'Working Hours' that is 9am-5pm, Monday to Friday, and then 'apply' this Time Period to the Host via a field called the 'Check Period', i.e. 'What period of time should I actively check this Host'.
Service Checks can be configured to have a fixed Time Period or to inherit the check period from the Host. This will apply to all Service Checks whether the Service Checks are applied individually via the 'Service Checks' tab or in bulk via the addition of a Host template.
Another usage of Time Periods within a Service Check context is that they allow the arguments of a Service Check to be modified during a certain Time Period. For example, if backups occur every evening between 1:00 am and 4:00 am then the load average on the file servers will be higher than usual, and may trigger alerts and false errors. To correctly account for this, an administrator could create a Time Period called 'backups' which is '01:00 ' 04:00, Monday ' Sunday' and then override the load average check to say:
Normal: check_nrpe -H $HOSTADDRESS$ -c check_load -a '-w 5,5,5 -c 9,9,9' Time Period 'backups': -H $HOSTADDRESS$ -c check_load -a '-w 9,9,9 'c 15,15,15' As can be seen, during the Time Period 'backups' the warning and critical levels are significantly higher ' meaning if the load average goes above '5' during the 'backups' Time Period then the Service Check stays in an 'OK' state and no alerts are sent. However if the load average goes above '5' AFTER the backups Time Period has elapsed, i.e. normal hours, then the Service Check will change to a 'WARNING' state and alerts may be sent, if the notifications profiles are configured as such.
In Opsview Monitor, by default, there are four Time Periods:
These Time Periods can be viewed by clicking on the 'Time Periods' menu option with the 'Settings' tab:
Once clicked, the Time Periods list page will load:
Within the list the four aforementioned Time Periods can be seen. To view the Time Periods details, i.e. the configuration window, the administrator can choose to either double-click on the relevant row or simply click 'Edit' within the contextual menu for the Time Period.
Once the edit window is opened, the administrator will be presented with a view similar to the one below:
There are seven days within the Time Period, Sunday through to Saturday. In each day, the hours can be defined in an 'HH:MM' format, and comma-separated for multiple ranges. For example, '00:00-24:00' means 'all day', '09:00-17:00' means '9am to 5pm', '00:00-09:00,17:00-23:59' means 'not 9-5pm', and so forth.
An important point to note is the hours defined do not go over the midnight boundary, for example "22:00-02:00" is not valid - instead use '22:00-23:59' on the first day, and '00:00-02:00' on the following day.