Hardware Model Normalization via Curated Hash Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managed networks face challenges in accurately representing hardware models due to inconsistencies in manufacturer names, product names, and model numbers across different discovery tools and manual entries, leading to reduced utility of configuration items in databases.
Innovation Solution
A normalization process is implemented to standardize hardware attributes by comparing entries in a database to a curated list of canonical values, using hash functions to match and update entries to consistent normalized values, and tracking normalization status.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple discovery tools and manual entry methods are used to populate hardware information, then the database can be populated with diverse hardware data, but the manufacturer names, product names, and model numbers become inconsistent
Solution Approach 1:
A normalization application is introduced as an intermediary component that sits between the discovery tools/manual entry processes and the CMDB. This application automatically standardizes hardware attributes by comparing them against a curated list of canonical values, thereby maintaining data consistency without reducing population efficiency
Solution Approach 2:
The system changes the state of hardware attributes from unnormalized diverse values to normalized canonical values. By applying normalization rules that transform variations like 'ABC Corporation', 'ABC Corp.', and 'ABC' into a single canonical form 'ABC Corporation', the system maintains consistency while preserving the ability to accept diverse input formats
2Device complexity
If hardware attributes are normalized to canonical values, then the number of hardware models per type is reduced to a manageable level, but the process requires traversing and comparing all hardware models in the database
Solution Approach 1:
A curated list of canonical hardware values is prepared in advance, organizing all possible manufacturer names, product names, and model numbers into standardized formats. This preliminary organization enables rapid matching during normalization without requiring complex real-time processing
Solution Approach 2:
Instead of processing all hardware models continuously, the system uses hash functions to create compact representations of hardware attributes and compares these hashes against pre-computed hashes of canonical values. This copying approach significantly reduces processing time while maintaining accurate matching
3Adaptability or versatility
If discovery tools use different criteria for forming hardware attributes, then various hardware representations can be captured, but the utility of discovered configuration items is reduced
Solution Approach 1:
The normalization application transforms hardware attributes by changing their parameter values from diverse unstandardized forms to canonical standardized forms. For example, it converts manufacturer name variations, product name variations, and model number variations into their respective canonical values, thereby maintaining flexibility in accepting diverse inputs while ensuring reliability of the standardized outputs
Data Source
AI summary
An embodiment involves persistent storage and one or more processors. The persistent storage may contain a plurality of hardware models specifying types of computing devices, wherein the hardware models include attributes representing manufacturer names, product names, and model numbers of the computing devices. The processors may be configured to: obtain, from the persistent storage, a hardware model; calculate a hash value over attributes of the hardware model; compare the hash value to hash values in a curated list, the curated list including normalized hardware models that map the hash values to normalized manufacturer names, normalized product names, or normalized model numbers; determine that the hash value matches a particular hash value for a normalized hardware model; and update, in the persistent storage, the hardware model to include at least one of a normalized manufacturer name, a normalized product name, or a normalized model number associated with the normalized hardware model.