Defining SLAs and Alerts

Click the New Alert button to create a new alert. The Add Alert pane displays a Type field. From the Type field’s pull-down list, select the alert type (Custom Expression, Exception, JMX Attribute, Log Regex, Low Memory, Server Down, Server Up, Thread Pool, URL Health).

Once you select the alert type, the pane expands to display additional fields to which you enter data describing the alert. Different alert types have different data fields. The Add Alert pane displays the appropriate fields for the alert type you selected and you then enter the required and optional data. (Note that most fields are required. If you omit data for a required field, the console highlights that field in red.) Where possible, the console supplies default values and you can change these or leave as is. You must also click Save when you have finished defining the alert conditions.

For example, the following figure shows the Add Alert pane for the Server Down alert type.

Enter a name for the alert and, optionally, a description. From the Severity pull-down list, select the severity level (Fatal, Critical, Major, Minor, Information) of the alert. For the Server or Group field, select the server or group to which the alert applies (the default is none), and then check if the alert is active.

A field for which data is required is marked with an exclamation point and highlighted in red if left blank. (The figure below has a field marked as required.) For more information on these fields, see below.

If you want to create your own custom alert, select Custom Expression and then complete the appropriate fields (such as the target object expression and type, the target value expression, threshold expression, batch size, among other fields) to define the alert condition.

Note, too, that the set of entry fields change depending on the type of the alert. For example, for a low memory type alert, you are prompted for the following data, as shown in the figure below. In addition, default values may appear in some fields depending on the type of alert. For low memory alert types, the severity field defaults to Critical. You can change the severity level. You can have the alert apply to any memory pool or select a specific memory pool from the list. The Threshold Ratio is also required (note that it is highlighted in red and marked with a red exclamation point) since this field is required for this type of alert.

The Memory Pool pull-down list includes all memory pools for the server. To see this yourself, click the Servers tab, then the Memory tab. (See the figure below.)

By contrast, for custom expression alert types, you are prompted for the following data, as shown below. Notice that for this type of alert, after you select the server, you can then use the Application field pull-down list to optionally select an application running on that server to which you want the alert the apply. (See Custom Alert Type Fields below for more information.)

The figure below shows a sample entry for an Exception alert type. Once you select Exception for the alert type, the screen expands and displays the additional fields of data you must enter. Both the Description and Application fields are optional, but you must select or enter data for the other fields. Click Save to save the new alert, or Cancel to quit without saving the alert. (See Exception Alert Type Fields below for more information.)

Click the alert name in the Manage Alerts summary pane. The pane displays the data previously entered for the alert, in the same format used for creating an alert, and you can make the desired changes.

Note that you cannot change the name of an alert. If you need to change the name of the alert, you should delete it and re-create it with the new name.

The sections below provide more information on how to set up alert fields.

The following sections explain the data you might select or enter for particular fields when setting up different types of alerts.

As explained previously, different alert types may prompt you for different data fields. Most alerts, regardless of type, include the following generic fields:

Note: Period and Batch Size are coupled together, since checks are performed only at poll time. That is, there is not a constant stream of updates.

Here’s a simple formula or rule of thumb for computing the Batch Size and Period values: (batch size * period) = (dampening period in milliseconds)

For example, here’s how you might apply this formula. Suppose you want to be alerted if the flow has been under high load for over one minute. You might model this as a Period value of 20,000 and a Batch Size value of 3.

Custom alerts are executed as Groovy script snippets. The execution context is exactly the same as for other scripts in Mule. The context bindings for these scripts are described in Script Context Bindings.

When creating a custom alert, you need to set the following fields:

You should also check the Pools tab in the Management Console for the actual pool names for your instances.

The example given for Threshold Expression is a very simplified example of a bundled Thread Pool Alert in the Management Console. You should feel free to explore the full power of using Groovy and Mule together.

Exception alerts allow you to intercept exceptions thrown in your Mule application.

When creating an exception alert, you need to set the following field:

There are a number of properties whose values you can incorporate into alert destinations and notifications. You incorporate these properties using the expression syntax shown above. Some properties are common to all alert types, while others pertain to specific alert types.

The following properties are common to all alert types. (The class com.mulesoft.console.alert.RaisedAlert contains the declaration of alert properties.)

Certain alert types have other accessible properties in addition to the common properties shown above. The table below shows the alert types for which additional properties are defined and the specific properties.