Home / crm-business-applications / ServiceNow Certified Implementation Specialist – Data Foundations (CMDB and CSDM) (CIS-DF) / Cheat Sheet
ServiceNow Certified Implementation Specialist – Data Foundations (CMDB and CSDM) (CIS-DF)

CIS-DF Cheat Sheet

CIS-DF Tests Whether CMDB Data Stays Trustworthy — Not How Many CIs You Discover

For implementers configuring ServiceNow CMDB and CSDM. Think out-of-box: identification and reconciliation rules, Service Graph Connectors, Data Manager policies and CSDM classes beat business rules and custom tables.

Check Your Readiness →
Among the harder certs
Avg: Not published by ServiceNow
Pass: 750 / 1000
Most candidates understand ServiceNow Certified Implementation Specialist – Data Foundations (CMDB and CSDM) (CIS-DF) concepts — and still fail. This exam tests how you apply knowledge under pressure.

CIS-DF Blueprint Map (CMDB and CSDM)

Weights are as quoted by secondary sources from blueprint KB0012913, which sits behind the Now Learning login. Most of the 75 multiple-choice and multiple-select items describe a data quality symptom and ask for the out-of-box feature that fixes it at the source with the least technical debt.

  1. 01
    Configuration (15%) — CI class hierarchy, identification and reconciliation rules, IRE, data source precedence, CMDB 360 multisource data
  2. 02
    Ingest (19%) — Discovery and the MID Server, Service Graph Connectors, IntegrationHub ETL, import sets, asset-CI model categories
  3. 03
    Govern (35%) — health KPIs, Data Manager attestation, certification and retirement, de-duplication, CI ownership, principal class filter
  4. 04
    Insight (20%) — CMDB Query Builder, saved and CMDB 360 queries, Unified Map, Data Foundations Dashboard and playbook
  5. 05
    CSDM Fundamentals (11%) — business capability and application, application service, technical and business service offerings, phased adoption

Wrong instinct vs correct approach

A team needs Jamf device data in the CMDB and proposes building a scripted REST integration
✕ Wrong instinct

Build the scripted REST integration with transform maps

✓ Correct approach

Install the Service Graph Connector for Jamf from the ServiceNow Store. It is built on IntegrationHub ETL, sends data through IRE and is maintained across upgrades, whereas custom code bypasses identification and adds technical debt.

Thousands of CIs have not updated in 90 days and leadership wants them removed
✕ Wrong instinct

Create a Data Manager delete policy for stale CIs

✓ Correct approach

Stale may only mean undiscoverable. Run attestation so owners confirm existence, then retire and archive, deleting last. Also check whether expired discovery credentials or schedules caused the staleness in the first place.

Duplicate server CIs reappear weeks after the team merged every one with the wizard
✕ Wrong instinct

Run the de-duplication wizard again and close the project

✓ Correct approach

Fix the cause: strengthen identification rules so serial number outranks name, or correct the source sending weak identifiers. For large repeating patterns, a de-duplication template resolves tasks automatically.

Management wants a weekly report of applications that depend on servers in one data center
✕ Wrong instinct

Filter the server list by data center and export it

✓ Correct approach

Build a CMDB Query Builder query across server, application service and application classes with multi-level relationships in the correct direction, then save it and schedule it as a report data source.

Know these cold

  • ▸ Identification rules decide create versus update; reconciliation rules decide who may write each field
  • ▸ IRE tries identifier entries in priority order; serial number should outrank name for servers
  • ▸ Data source precedence — ower priority number wins; data refresh rules let stale values be overwritten
  • ▸ Attestation confirms existence, certification checks values; then retire, archive and finally delete
  • ▸ Correctness KPI combines duplicate, orphan and stale metrics; completeness only counts empty fields
  • ▸ Extend the nearest parent class, never cmdb_ci directly, to keep inherited rules and patterns
  • ▸ Copy and extend an out-of-box Discovery pattern; editing the base pattern blocks its upgrades
  • ▸ Principal class filter limits CI choice on task forms without removing read access
  • ▸ Database hosting consumed by IT teams is a technical service; employee email is a business service
  • ▸ CSDM adoption is phased crawl, walk, run, fly, starting with foundation data

Can you answer these without checking your notes?

In this scenario: "A team needs Jamf device data in the CMDB and proposes building a scripted REST integration" — what should you do first?
Install the Service Graph Connector for Jamf from the ServiceNow Store. It is built on IntegrationHub ETL, sends data through IRE and is maintained across upgrades, whereas custom code bypasses identification and adds technical debt.
In this scenario: "Thousands of CIs have not updated in 90 days and leadership wants them removed" — what should you do first?
Stale may only mean undiscoverable. Run attestation so owners confirm existence, then retire and archive, deleting last. Also check whether expired discovery credentials or schedules caused the staleness in the first place.
In this scenario: "Duplicate server CIs reappear weeks after the team merged every one with the wizard" — what should you do first?
Fix the cause: strengthen identification rules so serial number outranks name, or correct the source sending weak identifiers. For large repeating patterns, a de-duplication template resolves tasks automatically.

Common Exam Mistakes — What candidates get wrong

Writing a business rule to stop one source overwriting another

Reconciliation rules already define which discovery source may update which attributes, data source precedence settles conflicts, and data refresh rules handle stale values. A scripted block bypasses IRE, adds technical debt and breaks when sources change.

Choosing a certification policy when the need is confirming a CI still exists

In Data Manager, attestation asks owners to confirm existence; certification asks them to verify specific values such as support group. Retirement follows attestation, then archive, and delete comes last because deletion loses history.

Assuming IRE merges duplicates or updates the newest matching CI

When one payload matches several CIs, IRE updates the oldest, flags the rest with duplicate_of and raises a de-duplication task. Nothing merges until someone resolves it with the wizard or a de-duplication template.

Expecting Discovery to populate owned by and support group

Discovery and Service Graph Connectors collect technical attributes. Organisational fields such as owned by, support group and business criticality come from authoritative imports, CI owners and certification policies, never from probes or patterns.

Relating a business application directly to its servers

CSDM keeps the business application logical. Each running deployment, such as production payroll and test payroll, is a separate application service, and only those carry infrastructure relationships. Claims processing would be a business capability.

Resources to help you pass ServiceNow CIS - Data Foundations on your first attempt

CIS-DF rewards the out-of-box fix at the source. Check whether your CMDB instincts avoid custom code.