Stress Settings

As a comprehensive real-time monitoring solution for multi-user environments, ControlUp displays a complex and flexible measure of system health, called Stress Level, for every monitored resource such as a folder, machine, user session or process. Stress Level is a numeric column, displayed in ControlUp’s grid with the following ranges: None, Low, Medium, High and Critical.

Use the Stress tab to configure the following alerts settings:

You can use the Stress Level column to quickly determine the health of your resources, for example by sorting the grid so that highly stressed resources are on top. The following exaplains how to configure the Stress Level column to optimally represent the current health status of resources in your environment.

Use Case

To learn how to configure stress settings to reduce IT ticket submissions, use our interactive use cases guide:

Stress Settings Tab Layout

Use the Stress tab in the Settings window to configure all Stress Level-related settings. As shown in the image above, this tab contains a folder tree identical to the tree displayed in the My Organization pane, a navigation bar to switch between resource types, a Counter selection area, an Applies To area to configure filters, a Settings area to configure the computation of the Stress Level value, and a Stress Levels area to configure the boundaries between different levels.

Stress Level Inheritance

Default Settings

By default, every ControlUp organization contains a single set of Stress Level settings, configured on the root folder of the organization. Unless configured otherwise, these settings are inherited by all child folders and the machines within them.

Subfolder Inheritance

ControlUp’s folder tree is designed to allow the user to arrange machines in folders according to their type. For example, you might want to separate your workstations from your servers, and further segment the servers folder into subfolders containing different types of machines. This type of arrangement is generally convenient, and is especially useful for configuring different Stress Level settings for different types of machines.

Filter Inheritance

Besides segmenting resources into subfolders, ControlUp also distinguishes between resources automatically, allowing to configure performance counter thresholds which are optimal for each monitored resource. This is done by using filters, which are pre-configured criteria configurable in the counter area.

Machine Filters

You can configure different thresholds for each machine type using the filter area of each counter configuration. ControlUp distinguishes between the following machine types:

General Purpose Server. A machine running a server-class Operating System, with no Terminal (Remote Desktop) Services installed. This could be a file server, an Exchange server, a Web server or an SQL server. These machines typically host a limited number of user sessions for administrative purposes, and have most of their resources consumed by background services.

RDS. A machine running a server-class Operating System with Terminal (Remote Desktop) Services role installed. These machines typically host multiple end-user sessions, running virtualized applications or full-desktop environments.

Workstation. A machine running a client Operating System, such as Windows 7, 8 or XP. This type of machine typically hosts a single user session with foreground processes (applications) consuming most of the machine’s resources.
By default, all filters within every counter inherit its default thresholds. By clicking on the filter name on the left, you can customize the thresholds for each filter, as described below.

Configuring the Stress Level Computation

The counter area of the Stress Settings tab allows you to configure which metrics contribute to the computation of the Stress Level column for each resource in ControlUp.

Per-Counter Configurations

The counter area includes a row for every column included in each of ControlUp’s views (Folders, Machines, Sessions, Processes, Accounts and Executables). Please note that each view supports a different set of columns. You can switch between views using the navigation buttons on top of the grid.

Each counter row includes several settings which configure the contribution of that counter to the total Stress Level of the record.

Yellow and Red

Every counter has a Yellow and a Red zone, with configurable numeric boundaries. In the example above, a machine CPU column’s default settings are 80% for Yellow and 90% for Red. Once a machine’s CPU usage climbs to 85%, the cell in its CPU column will become yellow. If the CPU usage drops below 80% again, this cell will go back to green. These changes in the grid should be instantaneous.

Duration

Once a Yellow or a Red boundary is crossed, ControlUp tracks the time the value of the counter stays above that boundary. You can configure ControlUp to increase the resource’s Stress Level when this happens, specifying how long should the value stay above the threshold. For example, you may decide that if a machine’s Disk Queue Length value stays over 1 for 1 minute, this may indicate an I/O bottleneck and should affect the machine’s Stress Level, and if the value exceeds 2 for a minute it may indicate a severe I/O issue you might want to be displayed in red, as shown:

Load

The Load value determines how many points should be added to the value of the record’s Stress Level column when a threshold is crossed for the time duration described above. For example, in the Disk Queue example above, if the value stays between 1 and 2 for a minute, the Stress Level will be incremented by 1 point. If the value is above 2, the Stress Level will be incremented by 2 points.

Severity By

To change the value used by the information grid to display the performance data of a column and modify the cell color accordingly, select a computation method from the Severity by drop-down list. The following values are available:

• Current Value. The column displays the present point value of the counter. This is recommended for counters such as Memory Utilization or disk free space, for which knowing the most current present value is most valuable.
• Max. The maximum value recorded in the counter since its sampling started. Valuable mainly for peak analysis and capacity planning.
• Min. Same as above, referring to the minimum value.

• Max In History. The maximum value of the counter’s current buffer.
• Average. The average value computed on all values recorded by the counter since its sampling started. Valuable mainly for long-term analysis and establishing baselines.
• Average In History The average value of the counter’s current buffer. Valuable mainly for rapidly fluctuating counters, such as Page Faults/sec and CPU usage.

To illustrate the usage of the above values, consider the case of a machine’s CPU column. If you select Average In History in the Severity by drop-down list, you may witness a situation in which the counter will be colored red, while its displayed value is in the green range. The reason for this is the fact that the displayed value is based on the current value (e.g. 5%), while the severity color code is based upon the Average In History value, which may be high (e.g. 90%). This type of configuration makes sense in most environments, since a momentary peak of CPU usage is usually no cause for alarm, while a prolonged CPU load detected by the Average In History value my indicate a performance issue and justifies a color coded severity alert.

We recommend that you take extreme care when customizing the counter thresholds and their calculation sources. It is best to consider the variance and fluctuation rate of each counter when planning a change to these values.

N/A Color

Some counters have a complex computation mechanism, which may fail under certain conditions. For example, when a value of a performance counter cannot be retrieved. For each of the metrics collected by ControlUp, you may decide that a failure to collect a counter’s value in itself represents an issue and should change the color of the column to yellow or red. For example, the XenApp Load.

Configuring Stress Level Boundaries

You can customize the Stress Level boundaries s panel on the left side of the Stress Settings tab, you can customize the numeric boundaries between ControlUp’s stress levels.

By default, all resources in your organization inherit the following default stress levels boundaries:

• (No Stress) < 1
• 1 <= Low < 2
• 2 <= Medium < 3
• 3 <= High < 6
• 6 <= Critical

Just like the stress level computation settings, you can configure these limits on a folder basis. This means that a resource (machine, session or process) with a stress level of 7 may be considered Critical in one folder and Medium in another, according to the needs of your environment.

To customize the Stress Level boundaries for a subfolder:

1. Switch to the Stress Settings tab of the Settings Window.
2. Click on the desired subfolder in the organization tree.
3. Uncheck the “Default Configurations” checkbox in the “Stress Levels” panel just below the tree.
4. Adjust the numeric boundaries using the sliders or by typing the numbers into the fields corresponding to each level.
5. Click “Apply Settings” on the Home ribbon.

To reset default Stress Level boundaries for a subfolder:

1. Switch to the Stress Settings tab of the Settings Window.
2. Click on the desired subfolder in the organization tree.
3. Check the “Default Configurations” checkbox in the “Stress Levels” panel just below the tree.
4. Click “Apply Settings” on the Home ribbon.

Receiving Stress Level Alerts

You can configure ControlUp to alert you when resources in your environment reach a configured stress level. For more information, please refer to the Trigger Settings section.