CIS – Data Foundations (CMDB and CSDM) (CIS-DF)

Domain 1: Configuration — 15%

From an architectural standpoint, CMDB configuration defines the structure every ingestion and governance process depends on. You’ll work with CI classes, inheritance, attributes, relationships, principal classes, identification rules, reclassification, and multisource data.

The core mechanism driving this domain is consistent modeling. A CI must represent the right kind of object, contain dependable identifiers, and relate to other CIs in a meaningful way. Don’t create a custom class just because the existing hierarchy feels unfamiliar. Evaluate its intended use, inherited fields, reporting requirements, and lifecycle ownership first. Strong configuration gives IRE and downstream tools dependable context. Weak configuration simply stores more uncertain data.

Domain 2: Ingest — 19%

Ingestion is a pipeline problem. Keep it modular: collect the data, normalize its structure, map it to the correct CI class, pass it through an IRE-aware path, and make failures visible for remediation.

You’ll compare Discovery, Service Mapping, Agent Client Collector, Service Graph Connectors, IntegrationHub ETL, and import-set approaches. Your main hurdle is managing data flow smoothly while preserving source identity and reconciliation rules. Don’t treat a successful transform as proof that the CMDB was updated correctly. Follow the payload through identification, source tracking, attribute precedence, and error handling. Multiple sources can describe the same CI, so duplicate prevention and governed updates must be built into the ingestion process.

Domain 3: Govern — 35%

This is where CMDB programs either become useful or turn into expensive record collections. In production, completeness, correctness, and compliance need owners, thresholds, remediation paths, and deadlines.

You’ll need to understand CMDB Health, duplicate management, stale CI handling, attestation, Data Manager policies, lifecycle rules, and relationship governance. Here’s where things actually break: nobody owns failing health metrics, policies retire active infrastructure, or teams improve dashboards without fixing the source processes creating the bad data. Don’t chase a perfect score by suppressing evidence. Find the source, correct the rule or feed, and confirm that the problem stops returning. The goal is controlled data quality over time, not a one-day cleanup exercise.

Domain 4: Insight — 20%

Insight questions test whether you can turn CMDB data into a defensible answer. Watch out for tools that look similar but solve different problems.

Query Builder identifies CI populations and relationship patterns. Dependency views visualize connected infrastructure. Unified Map provides broader service context, while dashboards summarize measurable conditions. The testers love to ask which tool best exposes an upstream dependency, a class-based population, a business-impact path, or a health trend. Your best clue is the requested output: a query result, visual topology, performance trend, or impact analysis. Don’t select the most feature-rich interface by default. Study how CMDB relationships and service context support troubleshooting, change assessment, risk decisions, and operational planning.

Domain 5: CSDM Fundamentals — 11%

CSDM gives the CMDB a shared service vocabulary and a practical adoption path. From an architectural standpoint, the model connects business capabilities, business applications, application services, technical services, service offerings, products, and supporting CIs so different teams can reason about the same enterprise.

You’ll need to understand the CSDM domains, principal relationships, and the staged adoption approach commonly described as Foundation, Crawl, Walk, Run, and Fly. Align records with actual ownership and service delivery rather than forcing every existing table into a diagram. Don’t model a level of maturity the organization can’t sustain. Start with trustworthy foundational data, establish accountable services, and expand into portfolio, operations, and value-focused use cases as the supporting processes become ready.