Introducing health tests
You can use health tests to monitor the health and performance of individual servers or for the entire organization or environment. Health tests detect problems by collecting and analyzing data or simulating user interaction with the server at regular intervals. Because UC Diagnostics is a monitoring tool, its functions only discover the servers in your organization and run health tests against the servers.
Each health test can run against individual servers or against groups of servers. Some tests, such as the Active Directory Account Lockout Health Test, cannot be run against individual servers and are run at the Exchange organization or Skype for Business/Lync environment level.
When you create a test from the Home Page, the entire organization is targeted by default. In a clustered environment, health tests run against the active node only. Health tests read information from nodes as a whole through a virtual server name.
By default, health tests are executed using the Diagnostic Services account. The Remote Registry Service is required for some health tests. Ensure this is running on your client. Only health tests that are relevant to available servers are shown when you right-click a server and select Detect. You can configure any test against a group.
All health tests use global settings such as notifications and alternate credentials. It is recommended you configure your global settings before you configure your health tests. For more information on configuring global settings, see Specifying the test configuration .
Types of health tests
The Management Console includes the following health tests:
By default, it is recommended that these tests be run every 30 minutes except for the server health tests. For more information about server health tests, see About server health tests
Creating and configuring health tests
You can perform the following functions with health tests:
Creating a health test
If you select a container for a health test, the test runs against all servers in the group at the time of execution. Typically, the test would be configured for all servers. If you add a new server, the test automatically runs against that server (without needing to be updated). You can configure a test to run only once or at regular intervals during a specified time period.
When creating a test, you can set the conditions, error thresholds, and warning thresholds. You can also specify the number of alarms to be skipped before notifications are sent. You might want to ignore notifications for occasional spikes, and only be notified if there are consecutive alarms.
For example, if your tests are configured to alert on errors only, you might decide to ignore notifications until a counter exceeds the error threshold four times consecutively. You would enter 3 in the Alarms to skip before sending notifications box. In this case, you are notified after the error threshold is exceeded four times in a row.
Once the test is created, it runs immediately.
The Server Health test is preconfigured with general defaults. Some tests have counters preconfigured to skip an alarm the first time it occurs before sending a notification. For a list of these counters, see Counters with Non-Zero Alarm Values .