Question 360
A ServiceNow configuration project has two acceptance conditions. The first issue is: A team configures a new custom table unrelated to cmdb_ci and expects CI Class Manager to administer it. The second is: Two sources insert separate server CIs because hostname alone is not a stable unique key. Which paired approach is suitable?
Explanation
Correct Answer
Both controls are required. For condition (1), use normal table-definition administration for the non-CMDB table rather than CI Class Manager. CI Class Manager is designed for classes in the CMDB hierarchy. Verify: The non-CMDB table is managed correctly and no phantom CI class appears. For condition (2), configure a CI-class-appropriate IRE identification rule using sufficiently stable identifiers, and compare insert-versus-match results with the Identification Simulator before sending data to production. IRE uses class identification criteria to match an existing CI or create a genuinely new one; hostname alone may not uniquely identify a multi-source server. Verify independently: Two independently sourced payloads for one verified server select the same CI, while a distinct server remains a separate CI.
Incorrect Answers
Part (2) is relevant, but the pair is incomplete. Part (1) answers a different need: A custom CMDB integration uses a direct table insert and bypasses existing IRE identification logic. It should instead use: Use normal table-definition administration for the non-CMDB table rather than CI Class Manager. Both required outcomes must be verified separately.
Part (1) answers a different need: A custom CMDB integration uses a direct table insert and bypasses existing IRE identification logic. It should instead use: Use normal table-definition administration for the non-CMDB table rather than CI Class Manager. Part (2) answers a different need: Cloud inventory and Discovery report different operating system versions for the same CI. It should instead use: Configure a CI-class-appropriate IRE identification rule using sufficiently stable identifiers, and compare insert-versus-match results with the Identification Simulator before sending data to production. Both required outcomes must be verified separately.
Part (1) is relevant, but the pair is incomplete. Part (2) answers a different need: Cloud inventory and Discovery report different operating system versions for the same CI. It should instead use: Configure a CI-class-appropriate IRE identification rule using sufficiently stable identifiers, and compare insert-versus-match results with the Identification Simulator before sending data to production. Both required outcomes must be verified separately.