CIS-DF--Day10- CI Class Manager

Описание: 1. CI Class Manager Overview
CI Class Manager is a centralized interface that provides visibility and control over:

CI classes
Class hierarchy and relationships
CI attributes
Managed-by groups
Extensibility and model configuration
Identification & Reconciliation rules

Only users with admin or sn_cmdb_admin roles can create CI classes.

2. Class Hierarchy & Managed By Group

CI classes extend from parent classes (example: cmdb_ci → hardware → computer → server → Windows server).
Setting Managed By Group at class level cascades to all CIs under that class.
However, if a CI is associated with a Technical Service Offering, the managed-by group from the offering overrides class-level settings.


3. Technical Service Offering (TSO)
A technical service offering represents services managed by infrastructure teams (e.g., DB management, network, servers).
Characteristics:

Many-to-many relationship with business services
Not visible to customers directly
Represents underlying technical capabilities
Can store managed-by, support, change groups
Helps define ownership of infrastructure-level components

It is differentiated in the service offering table using the Service Classification field:

Business service offering
Technical service offering
Application service offering


4. Product Models & CMDB

Default product models can be set on CI classes but are optional.
Product models are part of Asset Management but visible in baseline.
Best used when a class consistently uses a single hardware/software model.


5. CI Attributes
Every CI class contains multiple attributes inherited through the class hierarchy.
In Class Manager, you can view:

Attribute names
Type
Reference
Length
Mandatory flags
Derived (inherited) vs custom attributes

New attributes can only be added at the relevant class, not inherited levels.

6. Identification & Reconciliation Engine (IRE)
IRE ensures data quality during imports or discovery.
IRE Components


Identification
Determines whether an incoming CI already exists or is new using key attributes (e.g., serial number + hostname).
Similar to determining if two objects are duplicates based on specific characteristics.


Reconciliation
Defines which data source has authority (precedence rules) to update specific attributes.


Deduplication
Identifies duplicate CIs and creates tasks for resolution.
Removing duplicates improves CMDB health scores.


Reclassification
Allows CIs to move:

Up (parent class)
Down (child class)
Sideways (peer class switch)




7. Coalesce vs Identification Rule

Coalesce is used only during manual imports (import sets) for insert/update logic.
Identification rules work for discovery and any CI ingestion and determine uniqueness.

They serve different purposes.

8. CMDB Health Considerations
Removing duplicates, correcting class mismatches, and ensuring proper ownership configurations directly improves CMDB health KPIs such as:

Completeness
Correctness
Compliance


9. Conceptual Examples Used in the Session
Atul illustrated CMDB concepts using relatable analogies:

Family roles for managed by groups
Birthday gifts for duplicate detection
Fruits for attribute comparison
These help contextualize identification rules and class hierarchy.