Discovery Pattern Extension vs Reclassification for Storage Switch Records

[Javeriaj]

Hi All,

I'm looking for guidance on the recommended approach for handling a Discovery classification scenario.

Discovery populating records into cmdb_ci_storage_switch. A subset of these discovered records can be identified by a naming convention (for example, names beginning with a prefix such as TRP_). These records are not actually storage switch from a CMDB modeling perspective and would be more appropriately represented as Initiator Groups.

The logic is straightforward:

• Discovery inserts a CI into cmdb_ci_storage_switch

• If the CI name starts with a specific prefix (e.g., TRP_)

• The CI should be created as an Initiator Group rather than a Storage Switch

Initially, I tested this by modifying the OOB Discovery pattern and was able to achieve the desired result. However, I recognize that modifying the base pattern is not upgrade-safe, so I'm trying to implement the solution using a Pattern Extension instead.

So far I've experimented with:

• Pattern Extension logic

• Filter conditions

• Insert Data operations

However, I haven't been able to get the CI created in the target class as expected.

My questions are:

1. Is a Pattern Extension the recommended approach for this type of class correction during Discovery?

2. Would a Class Switch/Reclassification mechanism be more appropriate?

3. Should Reconciliation Rules play any role in this scenario, or are they intended for a different purpose?

4. Would a Business Rule that reclassifies the CI after insertion be considered an acceptable approach, or is that generally discouraged for Discovery-created CIs?

5. Has anyone successfully redirected Discovery output from cmdb_ci_storage_switch to another CI class based on a naming pattern or attribute value? If so, what approach worked best?

I'm mainly looking for a solution that remains upgrade-safe and aligns with CMDB/Discovery best practices.

Thanks in advance for any recommendations or examples.