Creating a new or modifying an existing analytics configuration

In order to generate analytics output, whether seasonal events, related event groups, or event groups based on patterns, you must create a new analytics configuration or modify an existing analytics configuration. You choose the analytics type (related events, seasonal events, or both) you want to run during the create new analytics configuration operation.

1. Start the Configure Analytics portlet. See Viewing current analytics configurations.
2. Select the Create New Configuration icon to create a new analytics configuration, or highlight an existing analytics configuration and select the Modify Selected Configuration icon to modify an existing analytics configuration. The UI displays a dialog box that contains parameter fields for the new or existing analytics configuration. The dialog box consists of the following four tabs:

   Table 1. Tabs in the Create configuration window

   Tab	Purpose	Go to
   General	Specify general parameters, such as date range and filters to apply.	Step 3
   Related Events	Specify parameters that govern how the configuration generates related events.	Step 4
   Patterns	Specify parameters that govern how the configuration generates event patterns.	Step 5
   Advanced	Override the default Event identity to be used by this configuration, and use in generating seasonality, related events, and event patterns.	Step 6

3. Populate the parameter fields in the General tab of the dialog box with the details applicable to the analytics configuration.

   Name

   Enter the name of the analytics configuration. The name should reflect the type of analytics configuration you are creating.

   For example, TestSeasonality1 and TestRelatedEvents1 might be names you assign to analytics configurations for seasonality events and related events. The name for an analytics configuration must be unique and not contain certain invalid characters.

   The invalid character list is the list of characters listed in the webgui_home/etc/illegalChar.prop file.

   Description

   Type a description of this analytics configuration.

   Analytics Type

   Select Seasonal event analytics, Related event analytics, or both.

   Note: If you select Related event analytics then event pattern processing will automatically be performed on your data. Pattern processing runs against all specified event types: the default event type, and any additional event types that were specified when the system was set up. If you want to limit the additional event types used in pattern processing for this configuration, then you can do this in the Patterns tab, as described in Step 5.

   Date Range

   Select either RelativeFixed date range or FixedRelative date range

   Relative: Enter the time frame that is to be included in the analytics configuration. The relative time frame is measured in Months, Weeks, or Days.

   Fixed: The Start date and End date parameter fields are active. Enter an inclusive Start date and End date for the analytics configuration.

   Fixed date range: The Start date and End date parameter fields are active. Enter an inclusive Start date and End date for the analytics configuration.

   Relative date range: Enter the time frame that is to be included in the analytics configuration.

   Run every

   To schedule the analytics configuration to run at specific time intervals, enter how frequently the configuration is to run. When you enter a value greater than zero, the analytics configuration becomes a scheduled configuration.

   Note: This option applies to the relative date range only. You cannot apply this option to the fixed date range.

   Filter

   Formulate a filter to apply to the analytics configuration. It filters out events from the historical event database so that these events are not used as input for this configuration's analytics processing. For example, enter Summary NOT LIKE '%maintenance%'. The purpose of this filter is to exclude events related to maintenance periods from the analytics processing.

4. Populate the parameter fields in the Related Events tab of the dialog box with the details applicable to the analytics configuration.
   By default, related event groups are processed using the default event identity field(s) that is defined in the Advanced tab, as described in Step 6. If event patterns are enabled then the following conditions also apply.

   • In addition to the default event identity field(s), related event groups are also processed together with any additional event types that were defined in the Event Analytics Configuration wizard, as described in step 2 of Configuring event pattern processing.
   • If additional event types were defined in the Event Analytics Configuration wizard, then you can limit which of those additional event types are used with this configuration. This is done in the Patterns tab, as described in Step 5.

   Relationship Profile

   Select the strength of the relationship between the events in an analytics configuration. If this value is set to STRONG, there is more confidence in the result and a smaller number of groups produced.

   Automatically deploy discovered related event groups

   Select this option to automatically deploy related event groups that are discovered by this analytics configuration.

5. Configure event pattern generation settings in the Patterns tab.
   By default pattern capability is enabled. This means that event patterns are automatically created when the configuration is processed using the process described in Suggested patterns.
   When you create an analytics configuration, the fields that were defined for event pattern processing in the Event Analytics Configuration wizard are processed in the following way:

   1. If there are no additional event types then all events in the Historical Event Database that match the configuration filter and date criteria in your analytics configuration are considered when looking for groups. The Default event identity field(s), specified in the wizard, are used to pinpoint individual events. The Default event type field is used for the event type of the discovered related events, and is used to determine the event pattern.
   2. If there is one additional event type then all events that match the filter of that additional event type in the Historical Event Database (and also match the filter and date criteria in your analytics configuration) are considered first when looking for groups. The Event identity field(s) of this additional event type is used to pinpoint individual events. The Event type field of this additional event type is used for the event type of the discovered related events. All other records in the Historical Event Database (that is, those record that do not match the filter of the additional event type but which do match the filter and date criteria in your analytics configuration) are processed using the Default event identity field(s) and Default event type field, specified in item a.
      Note: The additional event type takes precedence over the default settings. Events which match the filter of the additional event type are processed first and then the remainder are processed using the default fields.
   3. If there is more than one additional event type then they are processed in order; that is, events that match the filter for the first additional event type are processed first, then the second, and so on. All remaining events that do not qualify for any additional event type filters are processed using the default fields. The order of additional types is configurable using this wizard.

   Note: By default, all additional event types apply to all analytics configurations. However, this can be changed by limiting additional event types, as described in this step.

   Enable patterns

   By default this option is selected. The following options are specific to event patterns and therefore are only available if Enable patterns is selected.

   Limit additional event types

   By default, patterns are processed using the default event type and event identity fields, together with all additional event types that were defined in the Event Analytics Configuration wizard, as described in step 2 of Configuring event pattern processing. Select this option to limit which of the additional event types are used with this configuration.

   The drop-down list displays the additional event types that are currently active. From this drop-down list, deselect any additional types that should not participate in pattern processing.

   Override default resource field(s)

   By default, patterns are also processed using the default resource field(s) that were defined in the Event Analytics Configuration wizard. Select this option to override the default resource field(s).

   From the drop-down list, select the resource fields to use for event pattern processing by this configuration.

   Filter

   Optionally specify an ObjectServer SQL filter to apply to the pattern. This filter restricts the events to which the pattern applies; for example, enter Summary NOT LIKE '%maintenance%'.

   Any suggested pattern, or pattern that is created based on the related event groups resulting from this configuration, will automatically have this filter replicated in the Events Pattern portlet, as described in Specifying event pattern criteria.

6. Populate the parameter field in the Advanced tab of the dialog box with the details applicable to the analytics configuration.
   Draft comment: DEIRDRELAWTON
   DL Oct 2019 RTC 69237 and 69238

   Override default event identity field(s)

   Select this option to enable the Event identity drop-down list.

   By default, seasonal events, related event groups, and event patterns, are processed using the default event identity field(s) that were defined in the Event Analytics Configuration wizard, as described in step 1 of Configuring event pattern processing. Select this option to change the default event identity field(s) for this configuration.

   Event identity

   From the Event identity drop-down list, select the database fields that identify a unique event in the database. Event seasonality runs on all events that are selected from the Event Identity drop-down list. For information about how to change the fields in the drop-down list, see Changing the choice of fields for the Event Identity.

7. Before running the configuration, take note of the following.
   Important: You must set the maximum Java heap size prior to running an analytics configuration that involves seasonality. To do that, follow the instructions in Set maximum Java heap size, then come back to this topic.
8. Click either Save to save the report without running, or click Save & Run to save and run the report. You can also cancel the operation by clicking Cancel.
   Note: Once you save the configuration, you can no longer change any settings in the Related events, Patterns, and Advanced tabs. At that point you can also not change the following parameter fields in the General tab:

   • General > Name
   • General > Analytics Type