Action / Alarm Settings


This section explains the Action and Alarm Settings that can be made in Applications Manager. To access the Action / Alarm Settings page, click the Settings tab and go to the Action / Alarm Settings link. You can change the following settings:

General

  • Enable Actions - When alarms are generated, actions are triggered for those alarms (if you have configured any). If you do not want the actions to be executed in spite of the alarms, deselect this option.
  • Limit RCA message of Health attribute by _ characters - to set a maximum character limit for the Root Cause Analysis (RCA) message related to the health attribute. This limitation prevents excessively long RCA messages that could affect performance, readability, or data storage.

Execute Actions Repeatedly

Using this functionality, you can trigger an action continuously during each poll until an alarm changes from critical or warning to clear, or you can limit the number of action executions for the same threshold. The three types of recursive actions involved are:

  • Availability- If a monitor is down, you can choose to do one of the following:
    • Execute actions repeatedly until availability is restored. You can also restrict the number of executions by setting a specific value (1 is the default, and leave the field blank for unrestricted executions).
    • Execute actions associated with the monitor's availability repeatedly until they are Acknowledged/Picked Up.

      Note: The pickup option for availability alerts can only be accessed from the widget interface.

  • Health - If the Health of the monitor is in a Critical or Warning state, you can choose to do one of the following:
    • Execute actions repeatedly till the Health is clear, if the Health is critical/warning. You can also restrict the executions to a certain number of times by setting a value (1 is the default value and blank for unrestricted executions).
    • Execute actions associated to the health of the monitor repeatedly, until it is Acknowledged/Picked Up.
  • Attribute Level - If the attribute status is critical/warning, you can execute actions repeatedly till the attribute status is clear.

EMail Settings

Under Email settings, you have the option to:

  • Send a Server Snapshot as inline attachment to an email if the health is Critical. You can also select the Content Transfer Encoding for email action from a drop-down menu.
  • Send one email when multiple Email actions are configured for an alarm.

Monitor Error Mail

  1. Under the Error Mail Monitoring Settings, you can choose to send emails in case of any fatal monitor error like data collection being performed. This mail will be sent to the email address specified in the 'admin' user account. If no email address is specified, the mail will be sent to the email address specified in the 'Mail Server' settings.
  2. Furthermore, you can specify the number of times the error has to occur before email is triggered in the Check for consecutive polls before sending error box.

Date Format

Under Date Format for $DATE replacable tag in Alarm settings, you can set the required date format in the Email alarms. It can set to either:

  • Day Mon dd HH:mm:ss IST yyyy (Default) - Thu Feb 22 12:02:40 IST 2007 or
  • MM/dd//yyyy HH:mm:ss - 02/22//2007 12:02:40. See Replaceable tags for usage of $Date.

SMS:

By default , the complete information that you configure while adding SMS Action is sent via an SMS. Some SMS service providers restrict the length of characters sent through the SMS. This could result in truncated message delivery. If you would want to send only information on the Monitor, Attribute, and its Severity, deselect this option. For E.g., "Health of JBoss Server is critical" will be the SMS format that is received. (This is in addition to the message provided when creating the SMS action.)

Consecutive Polls before Reporting an Error:

The Critical, Warning, and Clear alarms are generated based on attributes that you have configured. You have control over the alarms that are being generated. Simply specify the number times after which the alarm should actually be generated. It would eradicate false alarms. You can set the number of Consecutive Polls Count that should be checked before reporting:

  • A service is down or an attribute is in a critical state.
  • An attribute is in warning state.
  • A service is up or an attribute is in clear state.

If you want to Applications Manager to check 3 out of 5 Consecutive Polls before reporting that a system (or any Monitor) is down or an attribute is critical, specify the value as 3 in the first and 5 in the second text box. Similarly, you can change the remaining four text boxes (warning, Clear) also. If you have set the poll as 2 out of 5 times before reporting a service is up, for the first time of polling, the service will be shown as Unknown. Only on the second poll, if the service is running, the status would be shown as 'Up'. Changes made will be reflected across all the monitors.

Apart from Action / Alarm settings, Polls before reporting an error, can be configured for threshold and availability of individual monitors. Refer Configuring Consecutive Polls.