SLAs

Feature

Different Types of SLAs are available in Germain UX, and can run against any KPI whether the KPI is preconfigured or custom, Business or Technical.

Fact-Based SLA

SLA that evaluates individual data-point or fact.

Statistical SLA

Statistical SLA evaluates to a collection of data points or facts.

Statistical SLA can be: an Aggregate SLA, Color Percentile SLA, Percent Change SLA, Predictive SLA or Smart Insight SLA.

Aggregate SLA

Aggregate SLA evaluates a period of time and compares it to a constant threshold.

Color Percentile SLA

Color Percentile SLA evaluates a period of time and compares the GREEN percentile to a constant threshold.

Percent Change SLA

Percent Change SLA evaluates a period of time and compares it to a prior period.

Predictive SLA

Predictive SLA Creates predictions based on past data-points or historical trends.

Smart Insight SLA

To automatically identify trends that aren't considered normal. Smart Insight SLA Evaluates a Business Period of Time (e.g. Year-end, Quarter-end, Month-end, Monday, etc) and compares it to other Business Periods.

Example of Smart Insights

• Alert1: Disk Usage is predicted to exceed 90% in a week from now

• Alert2: User Errors Volume has increased by 22.7% this monday compared with last monday

• Alert3: Sales Forecast predicted to be 70% accurate by month-end

• Alert4: eCommerce visitors predicted to be 40% higher by month-end

Configuration

Via Wizard

Go to Germain Workspace > Left Menu > Wizard > SLA

Via SLA Screen

Go to Germain Workspace > Left Menu > Analytics > SLA

How-to Enable Predictive SLA for your KPI

• Log on to Germain UX > Left Menu > Wizards
  

  

• Select the KPI for which you want to configure a predictive SLA (selection is disabled if you came from KPI View)
  

  

• Select the “Predictive-Curve” SLA type
  

  

• Specify the details of the value and timeframe you want to predict
  

  

• Select the alert template and action(s) you want the system to execute if the previous SLA check fails
  

  

How to enable Smart Insights for your KPI

• Germain UX > Left Menu > Automation > Alerts > analytical-sla-alert

 

• Germain UX > Left Menu > Analytics > SLA > Smart Insights SLA - Day

 

In addition to receiving individual email alert whenever a kpi is exceeding SLA, you can receive a summary email once per day that summarizes all the kpi that have exceeded SLA that day.

Germain UX’ Smart Insight engine can be enabled for any of the KPIs that are active in your system. KPIs can be Technology, Process or User Behavior-specific.

Example: Smart Insights Summary Email that is sent once per day.

 

Attributes

FactBasedSLA Context

The context object available from a FactBasedSLA has the following structure:

• sla (Object)

• threshold (Object)

• kpi (Object)

• fact (Object)

• value (Double)

• resource (String)

• violation (boolean)

• violationCount (int)

More info on the structure of these objects is available below. For example, if you have a LocalActionExecutor that executes a command when an SLA is breached, you can reference the variables above inside your command or arguments follows:

${context.fact.system.hostname}

${context.kpi.name}

Statistical SLA Context

The context object available from a StatisticalSLA has the following structure:

• sla (Object)

• kpi (Object)

• value (Double)

• resource (String)

• values (Map<String, Double>)

• timestamp (Object)

• result (Object)

More info on the structure of these objects is available below. These variables can be accessed in the same way as above.

SLA Object

The above objects have the following structure:

FactBasedSLA:
                name (String)
                color (Object)
                constantValue (String) – this is the SLA threshold
AggregateSLA:
                operator (Object)
                thresholdValue (Double)
                periodCount (Integer)
                periodGranularity (Object)
                rank (Double)
ColorPercentileSLA:
                operator (Object)
                thresholdValue (Double)
                periodCount (Integer)
                periodGranularity (Object)
PercentChangeSLA:
                operator (Object)
                thresholdValue (Double)
                periodCount (Integer)
                periodGranularity (Object)
                baselineGranularity (Object)
PredictiveSLA:
                operator (Object)
                thresholdValue (Double)
                periodCount (Integer)
                periodGranularity (Object)
                dataGranularity (Object)
                extrapolateValue (boolean)

Threshold Object:

name (String)
color (Color)
thresholdValue (Double)
thresholdExpression (String)

KPI Object

Name (String)
Description (String)

Fact Object

see documentation

Result Object

CurrentInterval (Object)

start (Object)

end (Object)

baselineInterval (Object)

start (Object)

end (Object)

measure (String)

unit (String)