Knowledge Center > Tutorials

Properly configured Groups and Dependencies act as an Alert suppression system in ipMonitor. When a critical resource fails, ipMonitor limits Alerts to the Monitor(s) defined as a Dependency rather than triggering Alerts for every member Monitor in the Group.

Defining Dependency relationships:

• Minimizes the number of redundant Alerts.
• Helps isolate the root cause of the problem.
• Prevents configured Recovery Alerts from attempting to restart services and applications if such an action isn't required.

This tutorial illustrates two common set-up scenarios, and provides sample configuration settings to help you configure Groups and Dependencies that suit your network environment.

Click the PDF icon to open a printable PDF version of this tutorial. It includes all the information found here, including images from the ipMonitor software, diagrams, and links to additional resources.

Dependency relationships help channel all Alerts that would otherwise be sent out when a critical resource fails, through one Monitor. The statements below help explain how Dependencies work:

1. Dependency Monitors must reach the Down state for ipMonitor to disable alerting for Group Members.
2. Even if multiple Dependencies are configured, each Dependency is treated as an independent test and will affect all Group Members when Down. In other words, only one Dependency must be in a Down state in order for ipMonitor to disable alerting for all Group Members.
3. Once a Dependency Monitor reaches the Down state, Members that are currently in an Up or Warn state will not reach Down or Lost. Therefore, these Member Monitors can only vary their state between Up and Warn.
4. If a Group Member progressed to a Down state prior to any Dependency, Failure Alerts will be triggered for that Group Member.
5. When a Dependency Monitor reaches a Down state, Group Members currently Down will have all future Failure Alerts suppressed, but will remain Down. Recovery Alerts will be triggered for Member Monitors in a Down state when they return to the Up state.
6. When all Dependencies return to an Up state from a Down state, all alerting functionality is restored.

A system administrator may want to monitor a server using the following Monitors: Ping, Service, Disk Space and HTML/ASP. The example below illustrates this scenario.

To set up this configuration, a Group and a Profile must be created as follows:

1. Log in to the ipMonitor Administration web interface.
2. Click the Monitors menu option and Add the Service, Disk Space and HTML/ASP Monitors.
3. Add the Ping Monitor, ensuring that its timing parameters are set up to test more frequently than the other Monitors. For more information on recommended timing parameters, refer to the Effective Timing and Notification Parameters section of this tutorial.

1. Click the Groups menu option and create a new Group for the server.
2. Click on Add Dependencies and select the Ping Monitor.
3. Click on Add Members and select the Service, Disk Space and HTML/ASP Monitors.

4. Click OK to save the changes.

1. Click the Notifications menu option. Locate the Profile you want to use, or create a new one for this set of Monitors.

TIP If you've created a new Profile for this set of Monitors, you must also create a new Alert to be sent when ipMonitor detects a problem.

2. Click the Add Groups to Profile button and add the newly created Group.
3. Click the Add Monitors to Profile button and add the Ping Monitor that acts as the Dependency for the Group.

TIP The correct Profile configuration should contain only two entries in the Groups and Monitors to Alert for section. For the purposes of this tutorial, the example provided shows the 'Internal Server ADMINBASE2 ' Group (that contains the Service, Disk Space and HTML/ASP Monitors) and the 'ADMINBASE2 Server' Ping Monitor. Ensure that the Service, Disk Space and HTML/ASP Monitors are not listed individually within the Profile. To remove them if they have already been listed, select them by clicking the respective checkboxes beside the Monitor names and click the Remove Selected Groups / Monitors button.

4. Click OK to save the Profile.

A network administrator may want to monitor various aspects of a remote network location through a VPN connection, such as a web server and a SQL server located at a hosting facility. Resources to be monitored could include: the main gateway, the remote network's gateway, the web server, the SQL server and any critical services or physical aspects on the individual servers. The example below illustrates this scenario.

The overall dependency configuration process is similar to that described in the previous scenario, with a few key differences:

• Separate Groups and Profiles must be created for each server (as in the example shown above).
• Server Groups and Profiles must have three Dependencies defined: the main router gateway Monitor, the remote network gateway Monitor and the individual server ping Monitor.
• Any remaining Monitors must be added as Members of that server's Group (for example, the Service, Drive Space and HTML/ASP Monitors shown above test aspects of the web server, while the second Group contains Service, Drive Space and ADO Monitors to test aspects of the SQL server).
• Alerts must be created separately for each Profile.

In order for Dependencies to work as effectively as possible, timing parameters must be adjusted to ensure that the Dependency Monitor fails before the other members of the Group. Otherwise, Group Members may reach the Down state while the Dependency is still in an Up or Warn state. Additionally, notification control parameters specified for the Dependency Monitor must be less than or equal to the notification control values specified for member Monitors.

The following example illustrates this concept:

1. A Ping Monitor is set up as a Dependency for a Server Group. It's configured to test every 60 seconds regardless of its state. Each test lasts a maximum of 16 seconds. If two tests fail, an Alert is sent and the Ping Monitor progresses to the Down state.

2. Member Monitors are configured to test every 120 seconds regardless of their current state. Each test lasts a maximum of 60 seconds. If the above Ping Dependency Monitor has reached the Down state, Alerts from the failing Member Monitors will be suppressed. These Monitors will not be able to reach the Down state. Suppressed Member Monitors remain in the Warn state regardless of the amount of test failures.

For more information regarding Dependencies, refer to the Monitors section of ipMonitor's Context-Sensitive Help system, then select Groups and Dependencies. The Context-Sensitive Help can be accessed by clicking the Help link located in the top right corner of ipMonitor's Administration web interface.

For information on other features and concepts related to those discussed in this article, refer to the following ipMonitor resources:

< Back

Last Updated: October 6, 2005 | What did you think of this topic?