CIS-DF · Domain 1 of 5

CIS-DF Domain 1: Configuration

Configuration covers how the CMDB is shaped before data arrives: the class hierarchy in CI Class Manager, the Identification and Reconciliation Engine that decides whether a CI is new or existing, and CMDB 360 multisource data.

Exam weight estimate
15%
Official objectives estimate
3
Estimated questions estimate
~11
Suggested study time estimate
9–11 hrs

About 13 min read Reviewed October 2026

Overview

Domain 1 is about the structure and rules of the CMDB rather than the data itself. The three blueprint objectives ask when and how to use CI Class Manager, when and how to use the Identification and Reconciliation Engine (IRE), and when and how to configure CMDB 360, also called multisource CMDB.

In CI Class Manager you browse the hierarchy that starts at cmdb_ci, review or add class attributes, see which attributes a child class inherits, and maintain identification rules, reconciliation rules, health inclusion rules, suggested relationships and the principal class setting. IRE is the platform service every well-built ingestion path calls: identification rules decide which existing CI an incoming payload matches, and reconciliation rules decide which data source may update which attribute. CMDB 360 keeps a record of what every source reported for each CI, so you can compare sources and recalculate values without importing again.

How it's tested

Domain 1 questions describe a modelling or data-arrival problem and ask which configuration feature solves it. Typical stems involve a new device type, a second data source overwriting good values, or a need to see which tool reported an attribute. Precise feature names matter, because several options will sound like plausible CMDB tools.

  • Which class or attribute

    A new kind of infrastructure must be stored, and you choose between an existing class, a new child class or a custom table outside the CMDB.

  • Who wins the attribute

    Two sources report different values for the same field, and you pick the reconciliation rule or data source precedence setting that controls the result.

  • Which source said what

    A team needs to compare or audit values reported by each integration, which points to CMDB 360 multisource data and its queries.

Reasoning it through

A stem says SCCM and Discovery both update the same Windows servers, and the operating system version keeps flipping between two values. Options include deleting one integration, writing a business rule, adding a reconciliation rule, or creating a new class. A reconciliation rule naming the authoritative source for that attribute, with data refresh rules if needed, solves it without code or data loss.

Objectives 3

Official objectives from the ServiceNow outline, in outline order.

1.1 Identify how and when to use CI Class Manager (table, class attributes, hierarchy)Browse, extend and configure CMDB classes in one place.
What it means
CI Class Manager is the CMDB application for viewing and changing the class hierarchy. Each class is a table extending cmdb_ci, and the manager shows its attributes, inherited fields, identification rules, reconciliation rules, health settings, suggested relationships and whether it is a principal class.
What to understand
Child classes inherit every attribute of their parents, so add a field at the lowest class that needs it. Before creating a class, check whether an existing class such as Linux Server or Network Gear already fits; new classes should extend the closest parent so that inherited identification rules and Discovery patterns still apply. Class Manager also lets you hide or label attributes and set which fields count toward completeness.
How it shows up
A scenario introduces a new type of device and asks whether to extend an existing class, add an attribute or create a standalone table.
Common confusion
Creating a custom table outside cmdb_ci feels simpler, but it loses identification, health scoring and relationship support.
1.2 Identify when and how to use IREThe engine that matches incoming data to existing CIs.
What it means
The Identification and Reconciliation Engine receives a payload from any source, uses identification rules to find a matching CI or create a new one, and uses reconciliation rules to decide which source may update which attribute.
What to understand
Identification rules contain one or more identifier entries, each a set of criterion attributes tried in priority order, such as serial number before name. Independent classes are identified on their own attributes, while dependent classes such as a Tomcat instance also need the relationship to their host. Lookup rules use related tables like serial numbers or network adapters. If an entry finds more than one match, IRE by default updates the oldest CI, marks the others as duplicates and opens a de-duplication task.
How it shows up
Questions ask which component prevents a second import from creating duplicate servers, expecting identification rules evaluated by IRE.
Common confusion
Candidates mix identification (is this CI new?) with reconciliation (may this source change this field?).
1.3 Identify when and how to configure CMDB 360/multisource CMDBStore and compare what each source reported for a CI.
What it means
CMDB 360, built on multisource CMDB, keeps each data source's reported values for a CI in addition to the single reconciled value shown on the CI record.
What to understand
With multisource data you can see that Discovery reported 16 GB of memory while an SCCM connector reported 8 GB, decide which source should be authoritative, and recompute the CI values after changing reconciliation rules without re-running every integration. CMDB 360 queries in CMDB Query Builder report on gaps such as CIs seen by one source but not another, which helps find unmanaged or missing devices.
How it shows up
Expect a scenario where an auditor needs to know which integration supplied a value, or where coverage between two tools must be compared.
Common confusion
Thinking CMDB 360 is a separate database or a duplicate CMDB, when it is source-level history kept alongside the same CIs.

High-yield concepts

The terms, rules and formulas this domain turns into questions most often.

  • cmdb_ci base class

    The parent table that every configuration item class extends.

    Remember
    Fields defined on cmdb_ci, such as name, owned by and support group, are inherited by all CI classes.
    Tested as
    Explains why an attribute added high in the tree appears on every child class.
    Trap
    Adding a field to cmdb_ci when only one class needs it, which clutters every form.

  • Class inheritance

    Child classes receive parent attributes, identification rules and reconciliation rules unless they override them.

    Remember
    Override at the child only when the parent rule cannot identify that type reliably.
    Tested as
    Questions on why a new class already has rules point here.
    Trap
    Rebuilding parent rules on each child, which creates conflicting maintenance.

  • Identification rule

    A class-level rule listing identifier entries that IRE tries in priority order.

    Remember
    Each entry is a set of criterion attributes; the first entry with a unique match wins.
    Tested as
    Central to every duplicate-prevention scenario.
    Trap
    Picking name alone as the identifier, which is rarely unique.

  • Independent vs dependent identification

    Independent CIs are matched on their own attributes; dependent CIs also need a relationship to a parent CI.

    Remember
    An application instance is identified together with the host it runs on.
    Tested as
    Shows up when software or databases duplicate across hosts.
    Trap
    Treating a dependent CI as independent and merging instances from different servers.

  • Reconciliation rule

    Defines which discovery sources may update which attributes of a class.

    Remember
    Unauthorised sources cannot overwrite attributes that another source owns.
    Tested as
    The answer when two tools keep overwriting each other.
    Trap
    Solving overwrites with a business rule instead of the IRE feature.

  • Data source precedence

    Priority between authorised sources when several may update the same attribute.

    Remember
    A lower priority number wins; data refresh rules let a lower-priority source update stale values.
    Tested as
    Used in scenarios that need a fallback when the main source stops reporting.
    Trap
    Assuming the most recent update always wins.

  • Discovery source

    The value that tells IRE which tool sent the payload.

    Remember
    Custom integrations should register their own discovery source so reconciliation rules can refer to it.
    Tested as
    Questions on why reconciliation ignores a custom import often turn on this.
    Trap
    Leaving every import labelled as a generic source.

  • De-duplication task

    A task IRE creates when one payload matches more than one existing CI.

    Remember
    By default IRE updates the oldest match, flags the others with duplicate_of and leaves the task for someone to resolve.
    Tested as
    Links Configuration to the remediation objectives in Govern.
    Trap
    Assuming IRE merges or deletes the extra matches on its own.

  • CMDB 360 / multisource CMDB

    Stores per-source reported values for each CI alongside the reconciled values.

    Remember
    Lets you recompute CI values after rule changes without re-importing.
    Tested as
    Answers any stem about comparing or auditing sources.
    Trap
    Confusing it with an archive or a backup of the CMDB.

  • Suggested relationships

    Class-level definitions of which relationship types make sense between classes.

    Remember
    They guide manual relationship creation so users pick valid types.
    Tested as
    Appears in relationship-quality questions.
    Trap
    Assuming they create relationships automatically.

Side by side

Concepts candidates mix up, compared on the point that decides the answer.

Identification vs reconciliation

ConceptMeaningKey differenceCommon trap
Identification rules Decide whether a payload matches an existing CI Controls create vs update Using them to stop a source from changing a field
Reconciliation rules Decide which source may update an attribute Controls who can write which field Using them to stop duplicates
Data refresh rules Allow updates when the authoritative value is older than a set period Handles stale authoritative data Ignoring them when the main source is retired

Ways to store a new device type

ConceptMeaningKey differenceCommon trap
Existing class Use the closest out-of-box class Keeps patterns, rules and reports working Rejecting it for a small naming difference
New child class Extend the nearest parent in Class Manager Inherits parent rules and attributes Extending cmdb_ci directly when a closer parent exists
Custom non-CMDB table A table outside the cmdb_ci hierarchy Loses IRE, health and relationship features Choosing it because it seems quick

Exam traps

Each trap is an answer that looks right. Learn the clue that should change your mind.

  1. Script the overwrite fix

    What candidates pick
    A business rule that blocks one source from updating a field.
    Why it looks right
    It looks precise and quick to write.
    Why it's wrong
    Reconciliation rules already control attribute ownership by source, so a script adds technical debt and bypasses IRE logic.

    Clue to watch for Two named sources disagree on the same attribute.

  2. Name as identifier

    What candidates pick
    Identify servers by name only.
    Why it looks right
    Names are always populated and readable.
    Why it's wrong
    Names change and repeat across environments; serial number and other strong attributes should come first in the identifier entries.

    Clue to watch for The stem mentions renamed or reimaged machines.

  3. Newest match wins

    What candidates pick
    IRE updates the most recently created match.
    Why it looks right
    It sounds like sensible default behaviour.
    Why it's wrong
    By default IRE updates the oldest matching CI, marks the rest as duplicates and raises a de-duplication task; nothing is merged until someone resolves it.

    Clue to watch for Words like 'two existing records match'.

  4. Top-level class extension

    What candidates pick
    Extend cmdb_ci directly for a new storage appliance.
    Why it looks right
    It seems the cleanest place to start.
    Why it's wrong
    Extending the nearest parent keeps inherited attributes, rules and patterns that a cmdb_ci child would lack.

    Clue to watch for An obvious parent class already exists for the device family.

  5. CMDB 360 as backup

    What candidates pick
    Enable CMDB 360 to restore deleted CIs.
    Why it looks right
    Keeping extra data sounds like a recovery feature.
    Why it's wrong
    Multisource data records what each source reported; it is not a recovery or archive mechanism.

    Clue to watch for The requirement is to compare sources, not restore records.

Answer framework

Use these steps to answer Configuration questions about classes, IRE and multisource data.

  1. Decide whether the problem is structure (classes and attributes), arrival (identification), ownership (reconciliation) or visibility of sources (CMDB 360).
  2. For structure, prefer an existing class, then a child of the closest parent, and reject custom tables outside the CMDB hierarchy, since custom tables lose CMDB health, IRE and relationship support.
  3. For duplicates on arrival, look at identification rules, identifier entry order and whether the class is dependent on a host; independent classes rely on their own identifier entries instead.
  4. For overwrites between tools, choose reconciliation rules, data source precedence and data refresh rules before any scripting option.
  5. For questions about which tool reported a value or how sources compare, choose CMDB 360 multisource data and its queries.

Practice question

A company imports network switches from a spreadsheet and also discovers them with ServiceNow Discovery. The spreadsheet keeps overwriting the firmware version that Discovery collects. What should the CMDB administrator configure?

Want more like this? The CIS-DF practice test has questions for every domain.

Checklist

0 of 11

Tick each one you can do without notes. Your ticks stay in this browser.

Study strategy

Begin in a Personal Developer Instance with CI Class Manager open. Walk from cmdb_ci down to Computer, Server and Linux Server, and note which attributes and identification rules each level adds or inherits. Then open the Hardware identification rule and read its identifier entries in priority order.

Next, use the IRE simulation or a scripted call to the identification engine with two small payloads, one that matches cleanly and one that matches two records, and see what happens in each case. Add a reconciliation rule for one attribute and test that a lower-priority source can no longer overwrite it.

Candidates often skip CMDB 360 because it seems advanced. Read the product documentation on multisource CMDB and build one CMDB 360 query to see the per-source values.

Finally, write a one-page table: problem, feature, why. Expect this table to answer most Configuration items and to support the duplicate and ingestion questions in Domains 2 and 3.

FAQ

How much of the exam does the Configuration domain cover?

Configuration is weighted at 15% of a 75-question exam, which is about eleven questions as a planning estimate. ServiceNow does not publish per-domain question counts.

Do I need to memorise the class hierarchy?

Know the main branches under cmdb_ci, such as hardware, servers, network gear, applications and services, and how inheritance works. Exhaustive table names are less useful than knowing which parent to extend.

Is IRE tested here or in Ingest?

Both. Configuration asks how identification and reconciliation rules are set up, while Ingest asks which ingestion methods use IRE and why that matters for upgradeability.

What is the difference between CMDB 360 and multisource CMDB?

Multisource CMDB is the capability that stores per-source values; CMDB 360 is the name used for the experience and queries built on that data.

Can identification rules be different for child classes?

Yes. A child inherits its parent's rule by default, and you override it only when that type needs different identifiers.

Do I need scripting skills for this domain?

Very little. The exam favours configuration through Class Manager and IRE rules, and scripted workarounds are usually the wrong answer.

Other CRM & Business Applications certifications

Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.

Test your Domain 1 knowledge

You have covered class design, identification and reconciliation rules, and multisource data. Test whether you can pick the right feature in short scenarios.

Take the CIS-DF practice test Check your CIS-DF readiness

Keep going