2.2 Alarm Detection

An Alarm Detection configuration enables you to define criteria for generation of alarm messages. You select a condition, or combine a set of conditions, that within specific limits, generate an alarm message. To monitor the system alarms, you use the Web Interface. Note that  enables you to deliver alarm messages to SNMP monitoring systems, as well.

An Alarm can be in one of two states: new or closed. An open Alarm is an indication of a certain occurrence or situation that has not been resolved yet. A closed Alarm is a resolved indication.

To create a new Alarm Detection configuration, click the New Configuration button in the upper left part of the Desktop window, and then select Alarm Detection from the menu.

To open an existing Alarm Detection configuration, double-click on the configuration in the Configuration Navigator, or right-click a configuration and then select Open Configuration(s)... .

Alarm Detection Menus

The contents of the menus in the menu bar may change depending on which configuration type that has been opened in the currently displayed tab. Alarm Detection uses the standard menu items that are visible for all configurations, and these are described in the section Configuration Menus in 2.1 Menus and Buttons.

There is one menu item that is specific for Alarm Detection, and it is described in the following section, The Edit Menu.

The Edit Menu

Item

Description

Workflow Alarm Value Names...

To to define a variable to use in the APL code, see the APL Reference Guide, and the section below, Workflow Alarm Value for further information.

Alarm Detection Buttons

The contents of the button panel may change depending on which configuration type has been opened in the currently displayed tab. Alarm Detection uses the standard buttons that are visible for all configurations, and these are described in the section Configuration Buttons in 2.1 Menus and Buttons.

Defining an Alarm Detection

An Alarm Detection definition is made up of:

  • A condition, or a set of conditions, see the section below, Alarm Conditions

  • An object such as host, pico instance, or workflow, that the alarm should supervise

  • The parameter that you want the alarm to supervise, for example, Statistics value

  • Time and value limits of supervision

To create a valid alarm detection configuration make sure that:

  • The Alarm Detection includes at least one condition.

  • Two conditions within an alarm guard the same object: host, pico instance, or workflow.

  • Two conditions are set to the same time interval criteria.

To define an alarm:

Create an Alarm Detection configuration by clicking the New Configuration button in the upper left part of the Desktop window, and then selecting Alarm Detection from the menu.

The Alarm Detection

Click on the Edit menu and select the Validate option to check if your configuration is valid.

Click on the Edit menu and select the Workflow Alarm Value Names option to define a variable you can use in the APL code, see the APL Reference Guide, and in the Workflow Alarm condition, see the section below, Workflow Alarm Value.

Enter a statement that describes the Alarm Detection that you are defining in the Description field.

Select the importance priority that the alarm should have in the Severity drop-down list.

Use the Alarm Detection Enabled check box to turn alarm detection on or off.

At the bottom of the Alarm Detection configuration, click the Add button.

The Add Alarm Condition dialog box opens.

The Add Alarm Condition

Select a condition in the Alarm Condition drop-down list.

Alarm Conditions

The Alarm conditions enable you to define specific situations or events for which you want the system to produce an alarm. You configure a condition to produce an alarm whenever a certain behavior occurs, within specific limits.

Note!

  1. An alarm is generated only if ALL conditions in the Alarm Detection are met.
     

  2. The Alarm condition limits are reset:

    • Every time you restart the Platform

    • Every time you save the alarm configuration

    • When you resolve the alarm

    Alarms can be managed in the 1.2.3 Alarm Inspector Tab web interface.

The Alarm Conditions that you can choose from are:

  • Host Statistic Value

  • System Event

  • Pico Instance Statistic Value

  • Unreachable Execution Context

  • Workflow Alarm Value

  • Workflow Execution Time

  • Workflow Group Execution Time

  • Workflow Throughput

Host Statistic Value

The Host Statistic Value condition enables you to configure an alarm detection for the Host Statistic parameters. For further information see Host Statistics in 6.16 System Statistics.

The Host Statistic Value condition

ItemDescription

Host

Select a host server from the drop-down list

Statistic Value

Select from the drop-down list the parameter that you want the alarm to watch over. For detailed description of every Statistic Value, see 6.16 System Statistics.

Limits

Select a limit, either Exceeds or Falls below , upon which the alarm should be triggered. Check During Last to specify the time frame during which the Limits value should be compared. If a match is detected, an alarm is invoked.

Example - Configuring a Host Statistic Value condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.


You want the system to generate a warning if the primary host is being overworked.

  1. Configure an Alarm Detection with the Host Statistic Value condition.
     

  2. Select the statistic value Swapped in from Disk (blocks/s).
     

  3. Enter a limit of 1200 swaps-a-second, during the last 3 hours.
     

Configure an Alarm Detection



Configure the Alarm Condition

The alarm will be triggered only if the statistic value has been higher than 1200 throughout the last 3 hours. Note that if a momentary drop in value has occurred during the last 3 hours, the alarm will not be triggered.


System Event

The System Event condition enables you to setup an Alarm Detection for the various Event types.

The System Event condition

ItemDescription

Type

Select an event-related reason for an alarm to be invoked. For a detailed description of every event type, see 4.3. Event Types.

Filter

Use this table to define a filter of criteria for the alarm messages that you are interested in.

To define an entry, double-click on the row.

The Edit Match Value dialog box opens. Click the Add button to add a value.

Limits

Specify the condition for the alarm to be triggered. The options are based on the number and frequency of occurrence of the event: Occurred Once, Occurred More Than, Occurred Less Than. In During Last, specify the time frame during which the Limits value should be compared. If a match is detected, an alarm is invoked.

Example - Configuring a System Event condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.


  1. Configure an Alarm Detection that applies the System Event condition.

    Configure an Alarm Detection


  2. On the Edit Alarm Condition dialog box, from the Event Type drop-down list, select Workflow State Event.
     

  3. On the Filter table double-click workflowName; the Edit Match Value dialog box opens.
     

  4. Click Add to browse and look for the specific workflow.
     

  5. Enter a limit of occurred more than 3 times during the last 24 hours.

    Select an Alarm Condition

The alarm will be triggered by every 4th occurrence of a "Workflow State Event" during the last 24 hours.

Pico Instance Statistic Value

The Pico Instance Statistic Value condition enables you to configure an Alarm Detection that guards the pico instance statistic value of a specific EC. For further information about the Pico Instance, see 6.10 Pico Viewer.

The Pico Instance Statistic Value Condition

ItemDescription

Pico Instance

From the drop-down list select the pico Instance of which you want to collect statistical data.

Statistic Value

See Pico Instance in 6.16 System Statistics.

Example - Configuring a Pico Instance Statistic Value condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.

A telecom provider wants the system to generate an alarm if the following two events occur simultaneously:

  • The relevant pico instance (EC) memory is overloaded.

  • Too many files are open on that same particular pico instance.

 

Configure an Alarm Detection
 

  1. Configure an Alarm Detection that supervises EC1 with the Pico Instance Statistic Value condition. Use this condition twice:

    • With the Used Memory statistic value

    • With the Open Files Count statistic value

  2. Select the Alarm Condition Pico Instance Statistic Value.

    Select an Alarm Condition
     
  3. From the Statistic Value drop-down list, select Used Memory.
     

  4. Enter a limit of 900000 KB with - Note!- no time limit. This means that whenever this limit is exceeded, AND the other conditions are met, an alarm is generated.
     

  5. From the Alarm Detection dialog select the alarm condition Pico Instance Statistic Value once again.

    Select Another Alarm Condition


  6. This time use the statistic value Open Files Count.
     

  7. Enter a limit of 10000 files, without any time limit.

An alarm is triggered by every simultaneous occurrence of overloaded memory on EC1 AND too many open files, at any time.


Unreachable Execution Context

The Unreachable Execution Context condition enables you to configure an Alarm Detection that will alert you if the connection, between the platform and the EC that the alarm supervises, fails.

The Unreachable Execution Context Condition

ItemDescription

Pico Instance

See the section above, Pico Instance Statistic Value.

Note: Selecting Any from the drop-down list applies the condition to all the clients.

Unreachable due to normal shutdown

Check to invoke an alarm whenever the connection between the Platform and the client fails due to a normal shutdown of the client.

Example - Configuring an Unreachable EC condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.

A telecom provider wants the system to generate an alarm if connection to any EC cannot be re-established within 10 minutes.

  1. Configure an Alarm Detection that uses the Unreachable Execution Context condition.

    Configure an Alarm Detection

  2. From the Pico Instance drop-down list, select Any.

    Define the Alarm Condition
     
  3. Enter the time limit of During the last 10 minutes.

The alarm will be triggered whenever the system detects a loss of connection between the platform and one of its ECs, for a period that is longer than 10 minutes.

Workflow Alarm Value

The Workflow Alarm Value condition is a customizable alarm condition. It enables you to have the Alarm Detection watch over a variable that you create and assign through the APL code. To apply the Workflow Alarm Condition use the following guidelines:

  1. Create a variable.
     

  2. Assign the variable with a value.
     

  3. Setup the Workflow Alarm Value condition.


To Create a Variable name:

  1. From the Edit menu in the Alarm Detection configuration menu, select Workflow Alarm Value Names.The Workflow Alarm Value dialog box opens.
     

  2. Click the Add button and enter a variable name, e g CountBillingFiles.
     

  3. Click OK and then close the Workflow Alarm Value dialog box.

To Assign a Value to the Value Name:

In the APL code, include the command DispatchAlarmValue. For example:

consume { dispatchAlarmValue
("CountBillingFiles",1);
udrRoute(input);
}

To Configure the Workflow Alarm Value Condition:

  1.   At the bottom of the Alarm Detection configuration, click  Add ; the Add Alarm Condition dialog box opens.

  2.   From the Alarm Condition drop-down list select  Workflow Alarm Value .

  3.   From the Value  drop-down list, select the name of the variable that you created.

  4.   Click  Browse... to select the Workflow that the Alarm Detection should guard.

  5.   Configure the Limits according to the description of The Workflow Alarm Value and click  OK .

The Workflow Alarm Value configuration

ItemDescription

Value

Select an alarm value from the drop-down list.

Workflow

Click Browse... to enter the workflow instance(s) that you want to apply the alarm to.

Limits

Summation: Select this check box to add up the dispatchAlarmValue variable (countBillingFiles in the figure above, The Workflow Alarm Value configuration) whenever it is invoked. Alarm Detector compares this total value with the alarm limit (exceeds or falls below), and generates an alarm message accordingly.

Note: Selecting Summation means that the During Last entry refers to the time period during which a sum is added up. Once the set period has ended, that sum is compared with the limit value.

For All Workflows: Select this check box to add up the values (see Summation above) of all the workflows that the alarm supervises. Alarm Detector compares this total value with the alarm limit (exceeds or falls below), and generates an alarm message accordingly. Note: This check box can only be selected when Workflow is set to Any.

For further information about Limits see the section above, Host Statistic Value.

The Workflow Execution Time

The Workflow Execution Time condition enables you to generate an alarm whenever the execution time of a particular, or all workflows, exceed or fall below the time limit that you specify.

The Workflow Execution Time configuration

ItemDescription
WorkflowThe default workflow value is Any . Use this value when you want to apply the condition to all the Workflows. Otherwise, click Browse to select a Workflow that you apply the condition to.

Example - Configuring a Workflow Execution Time condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.

A telecom provider wants the system to identify a workflow that has recently run out of input, and to generate an alarm that warns about a processing time that is too short.

  1. Configure an Alarm Detection to use the Workflow Execution Time condition.

    Configure an Alarm Detection
  2. Click Browse...; the Workflow Instance Selection dialog box opens.
     

  3. At the bottom of the dialog box click Any.
     

  4. Set a limit of Falls below 2 seconds.

    Configure the Alarm Condition

An alarm is generated whenever an active workflow seems to process data too fast (in less than 2 seconds).

Workflow Group Execution Time

The Workflow Group Execution Time alarm condition enables you to generate an alarm whenever the execution time of a workflow group exceeds or falls below the time limit that you specify.

The Workflow Group Execution Time configuration

ItemDescription

Workflow Group

Click Browse... to enter the address of the workflow group to which you want to apply the alarm.



Example - Configuring a Workflow Group Execution Time condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.

You want the system to generate an alarm if a billing workflow group has been active longer than 3 hours.

  1. Configure an Alarm Detection that uses the Workflow Group Execution Time condition.

    Configure an Alarm Detection
  2. On the Edit Alarm Condition dialog box click Browse... to enter the workflow group you want the alarm detection to supervise.

    Configure the Alarm Condition
  3. Enter a limit of Exceeds 3 hours.

The alarm will be triggered if the workflow group has been active longer than 3 hours.

Workflow Throughput

The Workflow Throughput alarm condition enables you to create an alarm if the volume-per-time processing rate of a particular workflow exceeds, or falls below, the throughput limit that you specify.

The Workflow Throughput configuration

ItemDescription

Workflow

Select a workflow with the throughput value, the processing speed, that you want to supervise. For further information about the throughput value calculation, see Throughput Calculation in 3.1.8 Workflow Properties. An alarm is generated if the throughput value is not within the condition limits.

Limits

For information about Limits see the section above, Host Statistic Value.

Example - Configuring a Workflow Throughput condition

Note!

The parameters in the following example do not apply to any specific system and are only presented here to enhance understanding of the alarm condition.

You want the system to warn you on detection of decreased processing rate.

  1. Configure an Alarm Detection to use the workflow throughput condition.

    Configure an Alarm Detection
  2. On the Edit Alarm Condition dialog box click Browse... to select the workflow with the processing rate that you want to supervise.
     

  3. Enter a limit of Falls Below 50000 (batches, UDRs, Bytearray).

    Configure an Alarm Condition

The alarm will be triggered by every occurrence of a workflow slowing down its processing rate to a throughput that is lower than 50000 units per second.