System Analytics

Viewing Runtime Resources

In the Run-Time Resources tab, you monitor resource usage via one or more graphs. You can add multiple graphs and pin them as necessary to build a dashboard. For example, you can pin graphs for Tomcat CPU usage and Windows CPU usage to visually monitor the percentage of CPU consumption on the server while actions are performed by users at runtime.

Note the following:

• To add a graph, select an SBM server in the Host field that currently has System Analytics enabled in SBM Configurator. Next, select the component, group, and property that you want to monitor, and then click Add. The graph is added to the dashboard with various display options along the top.
• Click the pin icon to keep the graph on the dashboard page for the host server you selected. This ensures the same graph appears the next time you view the details for that host.
• Click the palette icon to change visual elements of the graph. For example, you might change the line color, enable the fill area and set the color, turn the grid view on or off, or set the graph to display using a fixed scale.
• Use the drop-down lists to add another graph or to filter the dashboard by component, group, or property if you have added multiple graphs.

Working with Properties

In the Properties tab, you enable property monitoring, create rules, and view a summary of properties in a tabular view.

For example, you can monitor the record count for both the Common Log (CL_LOG) and Event Log (EL_EVENT) tables in the database and have the system notify you when the count grows beyond an acceptable range. This helps avoid potential issues related to performance and disk space.

Note the following:

• Use the host, component, group, and property fields to filter the table and find properties that you want to monitor. Click Reset to remove the current filter and view the entire table again.
• To enable property monitoring, click the name of a property in the table. The following options are available:
  • Monitoring

    Enable or disable monitoring for this property. For example, if you change Windows CPU usage to Inactive, CPU usage for the Windows OS is no longer actively monitored and the last known usage percentage appears in any graphs that you might have added in Runtime Resources.

    In addition:

    • For active properties, specify a Polling interval to determine how often a property is checked. The default is 2000 milliseconds (2 seconds).
    • Select Use fixed scale in graphical views to used fixed minimum and maximum values on the y-axis for graphs that are created for this property. This enables you to easily view and compare graphs side-by-side. You can toggle this option on or off at run-time by accessing the display palette on a run-time resources graph.

  • Notifications

    Enable or disable the notification feature for this property. Administrators with access to SBM Configurator can specify a list of e-mail addresses in the System Analytics tab that will receive an e-mail when a failure occurs or a property range exceeds a threshold.

    Specify an interval to determine how often notifications are sent.

    • The default value is 0, which means a notification is sent each time a failure is recorded.
    • Change the value to -1 to send only one notification when the first failure occurs, but none after that.
    • Change the value to 1 to send the e-mail every hour, change it to 2 for every two hours, and so on.
  • Validation

    Enable or disable a validation rule for this property. Use the following options to configure your rule:

    • Range

      Determines the acceptable range (minimum and maximum thresholds) for a property. For example, set the range for Tomcat CPU usage to 0:90 to specify that any usage percentage below 90% as normal or within acceptable range. If you enable notifications for this property, any time Tomcat CPU usage exceeds 90%, a notification is sent to the list of addresses specified in SBM Configurator. In addition, if you are monitoring Tomcat CPU usage in a graph on the Runtime Resources tab, any spike in usage above 90% appears in red.

    • Period

      Specifies the monitoring period for this validation rule when the maximum range value is exceeded. For example, set the period to 60 to capture usage for 60 seconds when Tomcat CPU usage exceeds 90%. This settings also affects the time period that is displayed in the graphs that are available on the Failure History tab.

    • Failure percentage

      Set a percentage to apply to the maximum Range value for the specified Period. A failure is raised if the percentage is exceeded in that period. For example, enter 75 to raise a failure if 75% of the Tomcat CPU usage counts that are collected during the 60 second period exceed the 90% maximum threshold.

  • Save

    In some cases, you may need to reset monitoring for a property. For example:

    • If you set the notification interval to -1 and you fix the issue without restarting SBM Tomcat, you can click the Save button to reset the failure count that the system is tracking.
    • If you need to clear validation data that has been gathered before the validation period ends, edit the property, and then click Save to restart the validation period.

Viewing Failure History

In the Failure History tab, you can view the history of failures for properties that exceeded a configured threshold.

Note the following:

• Use the host, component, group, and property name fields to filter the list of failures in the table. Click Reset to remove the current filter and view the entire table again.
• Use the Start Date and End Date fields to find failures that occurred within a specific date range.
• Click Clear History to remove all entries or entries that meet certain criteria.
• The Validation Date displays the date and time when the failure was raised.
• A list of e-mail addresses that received a notification appears in the Notification Recipients column.
• Click view in the Chart column to view a graph of the time period when the failure occurred.