This documentation supports the 18.08 version of BMC Atrium Core.

To view the latest version, select the version from the Product version menu.

Configuring Key Performance Indicators for CMDB

The CMDB Dashboard is a visual indicator of the overall health of the CMDB at any given time. As a Configuration Manager, you can configure the Integrity and Completeness Key Performance Indicators (KPIs) to monitor the health of the CMDB. The CMDB Dashboard is shown in the following figure.



Key Performance Indicators

The following video gives you a quick overview of the Key Performance Indicators and how you can configure these KPIs for CMDB. 

  https://youtu.be/kxaTvW9T4mo

Integrity

The Integrity KPI captures the orphan CIs and duplicate CIs in a datasource. 

Duplicate CIs- Duplicate CIs in the CMDB data can lead to reconciliation failures, multi match errors, and incorrect data in consuming applications. The Reconciliation Engine identifies the Duplicate CIs as per the standard Reconciliation Engine identification rules.

Orphan CIs-  An Orphan CI is an instance of weak CI class that does not have any “Weak” relationships. Identifying the orphan CI instances is only applicable to a production dataset.

Completeness

You can use the Completeness KPI to evaluate if a CI has missing attributes. It is an indicator of the quality of data in the CMDB. 


To view the KPIs on CMDB Dashboard

On the CMDB Dashboard, you can use the tool-tips for Integrity and Completeness KPIs to view health results rather than having to drill down inside the Dashboard utility for information. 

The tool tip for the Integrity KPI shows the classes that have been excluded for calculating Integrity and the number of CIs included in each of those classes. Similarly, the tool tip for the Completeness KPI shows the classes and its associated CIs that were excluded for calculating completeness.

Under the Dataflow section, you can click Datasources to view the details of the classes and datasets that were excluded for Datasource utility processing.


To specify the classes to be excluded for processing KPIs and datasources

You can specify the classes that you want to exclude while processing CIs for Completeness and Integrity. You can also define classes and datasets that you want to exclude from being processed for the Datasources utility.

Perform the following steps to configure the exclusion parameters:

  1. On the CMDB UI, go to Configurations > Configure Dashboard Parameters > Configure Dashboard and Datasources. The Dashboard Configurations page opens.
  2. Select the Enable Duplicate Processing checkbox to enable the processing of Duplicate CIs.
  3. Specify the parameters that you want to exclude, as shown in the following tables:

    KPI exclusion parametersDatasources exclusion parameters
    Completeness: Exclude Class ProcessingExclude Class Processing 
    Duplicate: Exclude Class ProcessingExclude Dataset Processing
    Orphan: Exclude Class ProcessingExclude OverlayDataset Processing
  4. Save the configuration changes.

 For more information about the exclusion parameters, see  Excluding classes and datasets for processing Dashboard parameters

These changes take effect when the Dashboard Utility refreshes the next time. The Dashboard Utility refresh cycle can be scheduled on the CMDB UI from Configurations > Configure Dashboard Parameters > Configure Schedule.


To specify the attributes for Completeness

Perform the following steps to configure the completeness KPI.

  1. On the CMDB UI, go to Configurations > Configure Dashboard Parameters > Configure Completeness.
  2. On the left pane, filter on a class name and select a class for which you want to define the attributes for completeness.
    You can see the attributes for that class.
    You can change its priority. Two or more attributes can have the same priority.
  3. You can delete an attribute if you do not want to consider it for completeness. 
  4. You can add another attribute that you want to consider for Completeness.
    1. To add a new attribute, click Add New Configuration
    2. Select an attribute from the drop-down list and assign a priority to it.
  5. Save the configuration.

These changes take effect when you refresh the Dashboard Utility the next time. The Dashboard Utility refresh cycle can be scheduled on the CMDB UI from Configurations > Configure Dashboard Parameters > Configure Schedule.


Formulas for calculating KPIs

Originally, the completeness calculations were based on the Reconciliation Identity Rules. Now, a copy of the default identity rules is available under Dashboard Configurations that allows users to change the attributes by class as per their requirement. The classes that are excluded for processing are not considered in the KPI  calculations.  

You can get the count of Processed CIs and Reported CIs from the CMDB:Dashboard form.

The KPIs are calculated as per the following formulas:

1) IntegrityWithoutDuplicates% = [(Processed Orphans CIs - Reported Orphans CIs)/Processed Orphans CIs] X 100

Here, the Strong classes do not qualify to be part of the above formula. Hence, only the weak classes are considered for calculations.

2) IntegrityWithDuplicates % =

{(Processed Orphan CIs - Reported Orphan CIs)/Processed Orphans CIs X 100 + (Processed Duplicate CIs - Reported duplicate CIs)/ Processed Duplicates X 100}/ 2

3) Completeness % =  [(noOfTotalCIsProcessedForCompleteness- noOfIcompleteCIs)/(noOfTotalCIsProcessedForCompleteness)] * 100

4) Overall health %  = ( Integrity% + Completeness%)/2


Completeness Weightage Calculation

This section describes the calculation used to determine the completeness of a CI.

The Completeness calculation assigns the highest score to attributes that have Priority= 1 and the score decreases by 0.3% down the priority ranking.

If there are multiple attributes sharing the same ranking, the score is distributed equally among those attributes. The total score of a CI is a sum of the scores of all non-null attributes. A pre-decided range of scores decides whether the CI can be qualified as Complete or Incomplete. In this case, the criteria is 50%.  On the CMDB Dashboard, those CIs that are Complete are displayed in the Green band, whereas those CIs that are Incomplete are displayed in the Red band.

Example 1

To understand the calculation, let us consider class  BMC_Document that has attributes and its priorities assigned as follows:

  • Attribute1 with Priority = 1
  • Attribute2 with Priority = 2 

 Attribute1 is assigned a weightage of 50% because its Priority = 1.

Attribute2 gets a weightage which is 0.3 % lesser than Attribute1 because it has a Priority = 2.


As per the original Standard Identification rule, if Attribute1 has a missing value, then the CI is considered as Incomplete even though the Attribute2 with priority 2 has a value. 

Now, you can add new attributes to a class and define its priorities for the Completeness KPI.

Let us assume that we define two more attributes to this class BMC_Document, which are:

  • Attribute3  with Priority = 1 and 
  • Attribute4 with Priority = 2

When all the four attributes are populated, their weightage is as shown in the following table:

AttributePriorityWeightage
Attribute1 or Attribute 3150%
Attribute2 or Attribute4249.85%   [Its weightage is 0.3 % lesser than Attribute1 or Attribute3 because it has a Priority = 2]


Depending on which attributes have values, the completeness score is calculated as follows. 

Combination

Expected

Reported as Incomplete?

Reason

1

The value of Attribute1 is missing, but values of the other attributes are present.

Complete CI

No

Attribute3 is present which has weightage of 50%

If a Priority =1 attribute is present it takes a weightage of 50%. As the criteria for completeness criteria is satisfied, the CI is marked as complete. 

2

The value of Attribute3 is missing, but values of the other attributes are present.

Complete CI

No

Attribute1 is present which has weightage of 50%

If a Priority =1 attribute is present it takes a weightage of 50%. As the criteria for completeness criteria is satisfied, the CI is marked as complete. 

3

The values of Attribute1 and Attribute3 are missing, but values of the other two attributes are present.

Incomplete CI

Yes

It is an incomplete CI because both attributes of Priority =1 are missing.

4

The values of Attribute2 and Attribute4 are missing, but values of the other two attributes are present.

Complete CI

No

It is a complete CI because both attributes of Priority =1 are present.


Example 2

Let us consider another example of BMC_ComputerSystem class that has Serial No, Token Id, Hostname, and Domain attributes having priority 1. 

According to the completeness calculation, the weightage for attributes of priority 1 is 50%. When four attributes have priority 1,  each attribute is assigned a weightage of  12.5%.

Let's assume that we have four computer system CIs, Comp1, Comp2, Comp3, and Comp4 with attributes populated as follows:

CIAttributes populatedIs the CI complete?Reason
Comp1 Token Id attribute  onlyNoToken Id has a weightage of 12.5% which is less than 50% .
Comp2Token ID and Serial No attributes onlyNoToken ID and Serial No attributes have a weightage of 12.5% each, which totals to 25%, and is less than 50%.
Comp3Token ID, Serial No, and Hostname attributesNoEach of the attributes has a weightage of 12.5%, which totals to 37.5% and is less than 50%.
Comp4Token ID, Serial No, Hostname, and Domain attributesYesEach of the attributes has a weightage of 12.5%, which totals to 50%. As the criteria for completeness criteria is satisfied, the CI is marked as complete. 
Was this page helpful? Yes No Submitting... Thank you

Comments

  1. Craig Ridley

    Hi, "A pre-decided range of scores decides whether the CI can be qualified as Complete or Incomplete." Do instructions exist on how to configure the completeness threshold for the score range?

    May 13, 2019 08:37
    1. Maithili Deshpande

      Hi Craig, 

      The table under "Completeness Weightage Calculation" will guide with calculating completeness threshold - 

      When all the four attributes are populated, their weightage is as shown in the following table:

      Attribute

      Priority

      Weightage

      Attribute1 or Attribute 3150%
      Attribute2 or Attribute4249.85%   [Its weightage is 0.3 % lesser than Attribute1 or Attribute3 because it has a Priority = 2]


      If this doesn't help, please let me know if it is something specific that you are looking for. 

      Regards,

      Maithili

      May 23, 2019 07:25