Creating an alert template

Configuring alerts is an advanced feature. You must have a good knowledge of the RTM database and structure to be able to use this advanced RTM feature for generating generic alerts. Using a template, yuou can easily create alerts that are based on grid data. Create a new alert template in the Templates section of the Console tab.

1. Under the Console tab go to Templates > Alert.
2. Click Add on the upper right side.
3. Under the General tab, enter the required values and save the template to see the Data Source tab.
   You have to configure the following five tabs to create an alert template:

   • General: Define basic information such as name, severity, and type. After you save the template, the Data Source tab is displayed.
   • Data Source: Define how data is retrieved to evaluate if there is an Alert. A GUI builder can create expressions that are to be evaluated for threshold value.
   • Actions: Define email and syslog notifications.
   • Layout:: Define which columns are to be included in both the Alert Log and alert emails.
   • Breached Items: View the current breached alerts.
4. Go to the Data Source tab and complete the required fields and Save.
   You can choose one of the following expression types from the Data Type menu and click Save to see other options:

   • Table - Grid table names are listed from the database. The generic alert can pull data from any database table and create threshold values for it. Choose a table to base the alert and click Save to see other options.
     1. Create Data Source Metrics and Custom Data Input Items if you want to use them for building expressions in Data Source Items.
     2. Click Add in Data Source Items to build your expression.
     3. Click Save.

     The created expression displays on the Expression Items list. You can delete or edit the selected expression item. After you create an expression, you cannot change the database table that is associated with it.

     Restriction: Each alert can operate only on one database table. The Expression builder does not allow creating expressions that span more than one table. You can build a generic alert by associating an expression with it and by determining its threshold (Hi/Low values). You can also associate the alerts with your own custom scripts.
   • SQL Query - Use to define your own database query.

     If you are using SQL Query type Data Sources, then the Data Source returns rows of information about exceptional information in the cluster. When you define an Alert, you can select a column to aggregate within the table.

     • Click Check Syntax to verify expression or click Get SQL to check your SQL query.

       Either OK or Bad Syntax result is displayed. Refer to the Cacti log for the actual SQL error.

   • Script - You can create alarms with your own scripts. These scripts are useful when you need an alarm that requires information from more than one database table or some additional process or aggregation.
     If you are using Script type Data Source, then the Data Source returns two objects, a breach value and data to be processed by the Alert. So, for this type of Data Source, there is no aggregation available, the script must aggregate internally. With a Script Data Source, the data that returned can be:

     • CSV with a Header,
     • CSV without a Header,
     • HTML, or
     • Text

     The returned data is included in the Alert log, and any email sent to Administrators or users.

5. Go to the Actions tab and configure email notification and threshold settings.
   • If you are using expressions, the details for the Email Subject and Email Body contents are created by RTM.
   • The Notify Accounts can append or remove items from the two multi-select boxes while it specifies the accounts to be notified when the alert is breached. In Extra Alert Emails field, you can also specify extra email accounts that can receive alerts for this alarm.
6. Go to the Layout tab and define columns, sort order, and the direction. The Layout tab only exists for Expression and SQL Query types. For Script type you have to provide a script to define the layout. After the layout is defined, you can control the data to be returned in the report. For example, if you want to include the current pending reasons for jobs that are pending longer than a day, then add the pending reason column to the layout.
7. Go to the Breached Items tab and view all the breached items.