For Uptime, Transaction and Full Page Load monitors, Alert Rule is defined as the number of failures per location and the number of locations to fail from, as shown below for HTTP and Transaction monitors.
To define the Alert Rule, select the number of failures and the number of locations.
The User has set Alert upon 2 failures from 2 locations for some uptime monitor M.
The user will be alerted when M has failed:
– At least 2 times in a row from some Location 1.
– For the second time in a row from some Location 2 while M has not recovered from Location 1.
The table below illustrates the example.
For Server-Device, Application and Custom Monitors
For Server-Device, Application and Custom monitors, your system or application is being monitored locally on your machine by the installed Monitis Agent (see Downloading and Installing Monitis Smart Agent), or Monitis JMX Agent for Java/JMX monitor (see Downloading and Installing java/JMX Agent). Therefore, unlike Uptime, Transaction and Full Page Load monitors, there is no location testing for these monitor types.
For Server-Device, Application and Custom monitors, alert rule is defined based on the number of repetitive failures, as shown below for Bandwidth and Tomcat monitors.
For most Server-Device, Application and Custom monitors, you need also to specify the monitor’s failure condition as part of the Alert Rule.
The failure condition is generally crossing of the set threshold by some metric, as shown in the above screens for Bandwidth and Tomcat monitors.
To define the Alert Rule:
– Set the number of failures.
– Define the failure condition. Depending on the monitor type, the failure condition is defined:
- In the Alert Rule
- In the monitor’s settings
Defining failure condition from Alert Rule and monitor settings is described below.
Defining Failure Condition for Server-Device, Application and Custom Monitors
Defining Failure Condition from Alert Rule
For those Server-Device, Application and Custom monitors for which the failure condition is defined from Alert Rule, you will see the Monitor will fail when condition in the New Alert Rule window.
Select the metric you want to monitor from the combo box, and define the failure threshold for it.
The User has selected for some bandwidth monitor M:
– Monitor will fall when: output traffic is more than 512 Byte/sec.
The user will be alerted when:
At least 2 checks in a row yield in output traffic for M to be more than 512 Byte/sec.
Defining Failure Condition from Monitor’s Settings
There are monitors for whom the failure condition is defined not in the Alert Rule but in the monitor’s settings.
For such monitors (e.g. CPU, Memory, Drive, etc), the failure condition is defined in the monitor settings at the time of configuring the monitor, as shown in the below screen for CPU monitor.
The failure condition can be then changed at any time by clicking the icon on the top right corner of the monitor to open the context menu and selecting Settings.
A window will open showing your monitor’s settings.
Edit the metrics and thresholds and click OK.
Alert Rules for System Events Monitor
For System Events monitor, alert condition is defined not by failure (as a result of crossing of a threshold by the monitored metric), but by occurrence of some system event. Therefore, for System Events monitor there is no problem or recovery alert: alerts are sent simply when any of the events defined in the monitor’s settings takes place.
For Cloud Monitors
For Cloud monitors, you can set Alert Rule:
1. For the Cloud monitor itself to be notified when number of instances on your web server passes some threshold.
2. For cloud monitors, you can set up different types of uptime and server device monitoring (see Cloud Monitors): the added monitors will appear as separate monitors in your Monitis dashboard.
You can then specify alert rules separately for each individual monitor of your cloud monitor, e.g. HTTP and bandwidth as shown below.