Defining KPI Value as the First Operand

About this task

The specific steps for defining each of the two operands in a simple rule expression depend on your choice of operand types. The operand type determines the fields that appear in the Rule Expression panel on the Edit Rule Expression page. A specific set of fields is associated with each operand type, and whenever you change an operand type, the displayed fields change automatically. If you choose the KPI Value operand type for the first operand in a simple rule expression, Optimize displays the fields shown in the following figure and described in the subsequent task.

The following task assumes that you have already chosen KPI Value as the first operand in the Rule Expression panel on the Edit Rule Expression page.

To define KPI Value as the first operand in a simple rule expression

Procedure

In the Category list (in the Rule Expression panel on the Edit Rule Expression page), click the category of the KPI or KPI instance whose value you want to compare.

In addition to values supplied by categories that your system administrator created, the list could include the options described in the following table.

Category	Description
Analytic Engine	Provides data about the Analytic Engines, such as JVM memory consumed and thread count.
Intrinsic Process Metrics	Provides generic data about your processes as identified on the Problems page. For example, process cycle time (how long an instance takes to run) and step count (how many steps have completed successfully). Note: When creating a rule against the intrinsic process metrics Cycle Time by Process or Process Time against a Reading Value, the unit of measure is in milliseconds.
Additional options that might appear in the Category list:	GarbageCollectionType PersistenceQualifier Queue Router WebService Process names

Once you have chosen a category, the KPI list displays the names of all KPIs in the category you selected. Click the KPI whose value you want to compare.

In the Value list, click the value to use for the first operand of your simple rule expression. For a KPI Value operand, you can choose from the following three types of values: reading values, statistical values, or diagnosis values.

The following table describes the Data values.

Data Values
Value	Description	If chosen, second operand must be
Reading Value	Collected data value.	Number Math KPI Value (numeric)
Reading Maximum	Highest event value within a given collection interval for a KPI instance. This value type should not be used with Composite KPIs.	Number Math KPI Value (numeric)
Reading Minimum	Lowest event value within a given collection interval for a KPI instance. This value type should not be used with Composite KPIs.	Number Math KPI Value(numeric)
Reading Date	Date the KPI reading was collected (based on the time zone of the Analytic Engine).	Time KPI Value (date)
Reading Time	Time that the KPI reading occurred (based on the time zone of the Analytic Engine).	Number expressed using 24-hour clock KPI Value (time)
Reading Day	Day of the week that the KPI reading occurred (based on the time zone of the Analytic Engine).	Time KPI Value (day)
Reading Trend Up	KPI reading value or event value is higher than the previously collected value.	Number of consecutive KPI intervals.
Reading Trend Down	KPI reading value or event value is lower than the previously collected value.	Number of consecutive KPI intervals.

The following table describes the Statistical values.

Statistical Values
The values described in this table represent statistical information about Optimize data.
Value	Description	If chosen, second operand must be...
Statistic Mean	Historical mean for the KPI instance and the statistical interval.	Number Math KPI Value (numeric)
Statistic Maximum	Maximum historical value for the KPI instance and the statistical interval.	Number Math KPI Value (numeric)
Statistic Minimum	Minimum historical value for the KPI instance and the statistical interval.	Number Math KPI Value (numeric)
Statistic Stdev	Standard deviation for the KPI instance and the statistical interval.	Number Math KPI Value (numeric)
Statistic Confidence	An expression, represented by a percentage value between 0 and 1, of the confidence that readings are within one standard deviation of the statistical mean.	Number Math KPI Value (numeric)

The following table describes the Diagnosis values.

Diagnosis Values
The diagnosis values are based on trends that Optimize tracks automatically. Some diagnosis values evaluate against the number of consecutive KPI intervals or Event occurrences. For example, the expression Diagnosis Trending Up >= 3 would evaluate to “true” if the KPI reading value (or Event value) increased for 3 or more consecutive intervals. Note that interval based expressions (like all other expressions) also use the expression’s Intervals Before True setting to determine the number of intervals before firing a rule by combining the interval values (expressionIntervals + intervalsBeforeTrue – 1). For example, the expression Diagnosis Trending Up >= 3 with Intervals Before True set to “4” would evaluate to “true” if the KPI reading value (or Event value) increased for 6 or more consecutive intervals.” This is because intervalsBeforeTrue = 1 when the expressionIntervals = 3, intervalsBeforeTrue = 2 when the expressionIntervals = 4, intervalsBeforeTrue = 3 when the expressionIntervals = 5, and intervalsBeforeTrue = 4 when the expressionIntervals = 6 (when the rule fires). You can exclude the zero-value KPI readings from the calculations of the Diagnosis Above Mean, Diagnosis Below Mean, Diagnosis Above Normal, Diagnosis Above Normal, Diagnosis Slope Above Normal, and Diagnosis Slope Below Normal by setting the `stats.rules.calculations` property in the wrapper.config file to `false`.
Value	Description	If chosen, second operand must be...
Diagnosis Trend Up	KPI reading value or event value is higher than the previously collected value.	Number of consecutive KPI intervals.
Diagnosis Trend Down	KPI reading value or event value is lower than the previously collected value.	Number of consecutive KPI intervals.
Diagnosis Above Normal	KPI reading value is higher than the statistical mean plus the number of standard deviations specified in the rule expression. For example, if you have a rule using Diagnosis Above Normal with a value of 2, this part of the rule is satisfied when the reading value is more than two standard deviations above the statistical mean.	Number of consecutive KPI intervals.
Diagnosis Below Normal	KPI reading value is lower than the statistical mean plus the number of standard deviations specified in the rule expression. For example, if you have a rule using Diagnosis Below Normal with a value of 2, this part of the rule is satisfied when the reading value is more than two standard deviations below the statistical mean.	Number of consecutive KPI intervals.
Diagnosis Slope Above Normal	Rate of change between the last and previous KPI readings plus the number of standard deviations specified in the rule expression. For example, if you have a rule using Diagnosis Slope Above Normal with a value of 2, this part of the rule is satisfied when the reading value is more than two standard deviations above the statistical mean.	Number Math KPI Value (numeric)
Diagnosis Slope Below Normal	Rate of change between the last and previous KPI readings is below normal, plus the number of standard deviations specified in the rule expression. For example, if you have a rule using Diagnosis Slope Below Normal with a value of 2, this part of the rule is satisfied when the reading value is more than two standard deviations below the statistical mean.	Number Math KPI Value (numeric)
Diagnosis Above Mean	This is a count based value that is used in a rule expression with a specified number of standard deviations. It compares the number of successive readings above the historical mean to the average number of readings above the historical mean plus one standard deviation.	Ignored.
Diagnosis Below Mean	This is a count based value that is used in a rule expression with a specified number of standard deviations. It compares the number of successive readings below the historical mean to the average number of readings below the historical mean plus one standard deviation.	Ignored.

In the Comparator list, click the logical comparator (such as Equal To) to compare the two operands in the simple rule expression (also known as a Boolean expression).
In the second Operand list, select an appropriate operand and complete the remaining fields to finish defining the simple rule expression. If you select Number as the second operand, then in the related Value field, you would type the numeric constant that you want to compare to the first operand. If your first operand is of type KPI Value, for example, the unit of measure of your numeric constant must be the same as the unit of measure used by the selected KPI (shown to the right of the KPI's Value selection). For example, if the unit of measure used by the selected KPI is megabytes, Optimize interprets the literal in megabytes.
Click Save to save the simple rule expression.