Setting up a Workstation Alert for ConnectionStatus

Document ID : KB000010981
Last Modified Date : 14/02/2018
Show Technical Document Details
Introduction:

 There are 4 possible Connection States:

3 = disconnected

2 = connected, slowly or no data

1 = connected

0 = unmounted

 

Question:

 How can I build a threshold alert on a value of 0, unmounted?

 When a JVM is restarted, it jumps to 3. Then it returns to 1 when connected.

 When a JVM has been down for 30 minutes, it becomes "unmounted".

Environment:
All APM Releases
Answer:

Solution:

From the CA APM Workstation User's Guide:

Danger Threshold

Danger Thresholds specify when Simple Alerts are to be triggered.

You set Danger Thresholds in conjunction with the Comparison Operator.

To set the threshold for a Danger alert:

  1. In the Threshold field, enter the value that triggers a Danger alert. 

    The units in the Danger Threshold values correspond to the value used in the metric grouping. For example, if you are making a Simple Alert for Servlet Average Response Time, the value is milliseconds.

  2. Set the ratio of excessive periods that must be met for the alert to be triggered.
    To do this, you put one value in the Periods Over Threshold field and another in the Observed periods field. For example, if you enter 8 and 10, then the danger alert will be triggered only if the metric exceeds the danger threshold in 8 of 10 observed periods.

    In this case, you may set 2 simple alerts. On one of the alerts, set the threshold >1, and on the other one set the threshold < 0.

    As always, supply the number of periods over threshold and the number of observed periods.

    Consider setting a summary alert over these two simple alerts. The Summary Alert will allow you to monitor for a value of 0, unmounted.

    Summary Alerts

    A Summary Alert provides a way to show the status of multiple underlying
    Simple Alerts with one overall status.

    A Simple Alert has one of four states:
    not reporting
    green
    yellow
    red

Because the Summary Alert state is defined as the worst state among the Simple Alerts it contains, Summary Alerts have no explicit danger or caution thresholds or comparison expressions, as with Simple Alerts.

This is an example of the format of a Summary Alert notification message: 
4/13/04 12:31:45 PM PST

The Summary Alert "Application Health" is in the danger state due to:

SuperDomain/<Acme>|<SimpleAlertName1> is in danger 
SuperDomain/<Acme>|<SimpleAlertName2> is in caution 
SuperDomain/<Acme>|<SimpleAlertName3> is normal 
SuperDomain/<Acme>|<SimpleAlertName4> is not reporting