Governance rules and responsibilities
Determines, Who is responsible for the data, what rules apply and what to do about incorrect or outdated CIs.
The question is, “How do you keep your CMDB right?”SERVICENOW • CIS-DF • LOCAL DECK
CMDB & CSDM · English exam
There are no cards in this mode. Choose another topic or drop the search.
QUICK GLOSSARY
Determines, Who is responsible for the data, what rules apply and what to do about incorrect or outdated CIs.
The question is, “How do you keep your CMDB right?”The process through which data enters or updates the CMDB from Discovery, connectors, imports, and other sources.
The question is, “How does the data get into the CMDB?”Turns CMDB data into visibility, conclusions, and decisions: what is connected, where gaps exist, and what needs to be fixed.
The big question is, "What does the data show and what do we do next?"Remember: Ingest → Govern → Insight. Upload data → keep it correct → understand it and make a decision.
A picture in the head
Open the section when individual answers seem unrelated.
Think of CMDB as a catalog of a big city. Every real or logical object – server, application, network or service – receives a card. CI. But cards of different objects should have different fields, so they are arranged in classes.
Don't confuse: A class is not a label, but a table with inherited behavior. Principal Class highlights important classes and helps limit CI choices in operational tasks.
First the organization decides, What are the real objects you need to manage?. Suitable standard classes, fields and relationships are selected for them. Sources then find objects and, through IRE, create or update CIs. After that, Health checks whether the necessary fields are filled, whether there are duplicates and whether the connections are correct.
glide.class.switch.enabled Turned off, the cross-branch switch creates a manual check problem.u_cmdb_ci.cmdb_ci_app_server and cmdb_ci_appl linkage many-to-many.If the Installed Application is identified only by name, the two installations may merge into overloaded CI. Configuration file path helps to distinguish installations. First, fix the identification rule, then in Duplicate CI Remediation Wizard choose Master CI. The system initially prefers a record with a large number of related elements; the rest communicate via Duplicate of, and de-duplication template makes the logic repeatable.
glide.required.attribute.enabledRecommended Fields is configured in CI Class Manager → Health → Completeness.Object → CI → class → inherited fields and rules.
Think of the IRE as passport control. The source brings the data, but does not have to write it directly to the CMDB. First, the system determines which object it is talking about and whether the source has the right to change its fields.
Don't confuse: Identification selects a record; Reconciliation selects whose value can be written in a field.
Each download method has its own role, but the safe ending is one. Payload must pass through the IRE. Discovery horizontally searches for infrastructure over the network; Service Mapping builds service dependencies from top to bottom; ACC sends data from a light agent; Service Graph Connectors typically connect cloud and observability sources; IntegrationHub ETL and Import Sets convert external data.
Discovery was the first to build a server and write an OS. Later, a less trusted import sent a different OS, but also a new RAM. Identification finds the same CI. Reconciliation protects the OS from overwriting but allows RAM to be filled. Multisource CMDB stores both source-specific variants, CMDB 360 shows the discrepancy, and the main record remains the winning value.
glide.identification_engine.skip_duplicates. Multisource is enabled by the property glide.identification_engine.multisource_enabledsource-specific values are stored in cmdb_multisource_data.Source → IRE → find CI → check the right to update → save.
Imagine running the library. It is not enough to put books in a catalog: you need owners, rules for checking, corrections and a clear end to the life cycle.
Don't confuse: Managed by – who manages CI; Support group – who supports; Change group – who handles change. Offering-level groups can be more accurate than class-level groups.
Governance does not start with a Delete button, but with four questions: Why you need a CMDB, who owns the data, how success is measured, and who fixes the problem. Governance Charter links outcomes, KPI and critical success factors. The Configuration Control Board sets direction and priorities, and domain owners confirm data that the CMDB team cannot know.
life_cycle_stage more detailed life_cycle_stage_status. Life Cycle Mapping [life_cycle_mapping] Transforms legacy statuses into a new model and Life Cycle Control [life_cycle_control] Specifies the permissible Stage/Status for each class. After activating Lifecycle Sync, unclosed mappings are checked in Discrepancy Report.First, check if the servers are really irrelevant and why the source stopped updating them. Legitimate exceptions are excluded from a specific metric through the Health Inclusion Rule. For a real error, Remediation Rule runs a workflow and creates a Remediation Task. Once confirmed, the End of Life policy may be Retire CI, then Archive it for eventual recovery, or Delete without recovery. For the group of duplicates Main CI is offered on three grounds: More relationships, fresh Updated, older Created.
Assign the owner → set the rules → find the problem → fix → correctly complete the lifecycle.
Imagine the dashboard of the car. The data is already downloaded and managed; now you need to see the state of the system and understand where to look next.
Don't confuse: A high score proves the health of only those classes, rules and CIs that fall within the scope. If the result looks odd, first check scope, inclusion rules, and freshness of scheduled jobs.
Insight answers not with a single score, but with a sequence: Measure → narrow the problem → see links or sources → take action. CMDB Health measures the quality of a managed CI set; Data Foundations checks prescribed practices and offers playbooks; special tools help investigate the cause.
Query Builder adds Database node, filter Location = Seattle, and connects Incident from Non-CMDB Tables. Properties select the desired output fields. Before a scheduled query, the number of results is limited. If you only need a visual chain of providers/consumers around one CI, open Unified Map faster: it is a single modern view on top of Application Service maps and Dependency View.
glide.cmdb.query.nlq.activated. Sensitive non-CMDB tables are blacklisted; scheduled results are limited.Health measures quality; 360 compares sources; Query Builder asks; Unified Map shows connections.
Think of CSDM as a general subway map. This is not a separate product, but an agreement on which entities to use and how to connect them so that different ServiceNow products understand the same model.
Don't confuse: Business Application is not production CI. Service Instance was formerly known as Application Service; Technology Management Service/Offering was formerly known as Technical Service/Offering.
CSDM combines three perspectives on one reality: What the business plans is what really works and what the consumer gets. If you use prescribed out-of-box tables, references, lifecycle and relationships, Incident, Change, APM, Security and other products see the same picture.
Inventory Portfolio concept: Business Application. Inventory vX Production Europe Service Instance, which is selected in the Configuration Item field from Incident or Change. If a single change covers many CIs, they are collected in the Dynamic CI Group And the group itself is still pointed in the field. Configuration Item. Its infrastructure is collectively managed by the Technology Management Service Offering. The business consumer receives the Business Service Offering associated with this Service Instance.
Application plan → working instance → technical support → consumed business service.
This section is a dispatcher, not a new topic. In the scenario question, first find a noun to manage. It almost always leads to the right tool.
Red flags: Direct write in cmdb ci bypasses IRE; removing exceptions for the sake of a beautiful score destroys the correct data. Archive picks when recovery is needed; Delete when it should be impossible.
The phrase “show which source values differ” leads to CMDB 360. "Who depends on this CI?" Unified Map/dependency view. Build a stored query into multiple classes and Incident to Query Builder. Fix the next CSDM maturity gap Data Foundations Dashboard.
Govern + Insight + Ingest ≈ 74% published weight, repeated in Quizlet sets.
Find the object of the question → choose the tool that controls this object.