CMDB Data Normalization in practice

[twdykes]

What’s your most effective method for CMDB data normalization? We are coming over from Cherwell CMDB and will need to perform a good amount of normalization as part of the migration.

[tiagomacul]

Migrating from Cherwell to ServiceNow is a significant step that requires moving away from simple 'data storage' toward a governed data model. Since Cherwell often allows for more loosely defined structures, the most effective method for normalization in ServiceNow isn't just a technical script, but a structured framework.

The challenge of data normalization is a long-standing one in the industry. Looking back at the CMDBf (CMDB Federation) initiatives established by pioneers like IBM, BMC, and CA, the core goal was always to create a unified language for IT assets. ServiceNow has matured this concept by moving from simple federation to a strictly governed, hierarchical data model.

When migrating from a tool like Cherwell, the most effective method is to embrace this inherited structure (the 'ITIL DNA') that ServiceNow applies to its classes. For instance, the transition from a generic Computer to Server, and then specifically to Linux or Windows Server, is not just about organization—it’s about preparing the data for specific management tasks.

1. Leverage the Content Service for Global Standards The most sustainable approach is to utilize the ServiceNow Content Service. By opting into this, the platform automatically normalizes discovery data against a global library of Manufacturers and Products. This ensures that your 'Legacy Data' aligns with the same standards that the 'industry giants' helped define years ago.

2. Hierarchy-Driven Normalization Effective normalization follows the class hierarchy. By ensuring data is correctly classified at the Hardware level before it flows down to specific server types, you prevent the 'patchwork' CMDB effect. Using IntegrationHub ETL and the IRE (Identification and Reconciliation Engine) allows you to cleanse and normalize data before it hits the target tables, maintaining the integrity of this hierarchy.

3. From Federation to Single Source of Truth While the old CMDBf model focused on federating disparate sources, ServiceNow focuses on Consolidation. Implementing Normalization Data Services rules ensures that variations like 'Dell Inc.' and 'Dell' are unified. This aligns with Pillar 5 of Digital Transformation (Data

Governance): ensuring the CMDB remains a reliable 'Single Source of Truth'.

Strategic Insight: From an Enterprise Architecture perspective, normalization is a marathon, not a sprint. As defined in the Now Create methodology, the focus should be on Value-Oriented Implementation. If the foundation (the hardware classes) isn't normalized, every process above it—from Incident to Asset Management—will lack the necessary context to deliver real business value.

For those navigating these complex migrations and looking for a practical roadmap based on these architectural pillars, this guide provides a deep dive: ServiceNow Now Create: Practical Methodology