CMDB Confidence Scoring for Change Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In configurable managed environments, the Configuration Management Database (CMDB) often receives conflicting data from multiple point products, leading to inaccuracies and synchronization challenges, which can result in adverse consequences when changes are made based on outdated or incorrect information.
Innovation Solution
A method is introduced that involves acquiring information from the CMDB about configuration items (CIs) affected by a proposed change, computing a confidence level to assess the accuracy of this information, and allowing the change only if the confidence level meets a pre-specified minimum value, thereby reducing the risk of adverse consequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the CMDB accepts data from multiple point products to maintain comprehensive configuration information, then the quantity and completeness of configuration data is improved, but data accuracy and reliability deteriorate due to conflicts and synchronization delays
Solution Approach 1:
The patent introduces a confidence level computation mechanism as an intermediary layer between the CMDB data and change management decisions. This mediator assesses the reliability of configuration data by computing confidence levels based on multiple factors including data source credibility, synchronization status, and conflict resolution outcomes, thereby enabling informed decision-making despite data inconsistencies
Solution Approach 2:
The patent transforms the static CMDB data into a dynamic assessment by introducing confidence level parameters. These parameters change based on data freshness, source reliability, and conflict status, allowing the system to adaptively weigh different data sources and synchronize their reliability assessments
2Productivity
If changes are made based on CMDB information without verification, then change implementation speed is improved, but the risk of adverse consequences increases due to potential information inaccuracy
Solution Approach 1:
The patent performs preliminary assessment of CMDB data reliability through confidence level computation before changes are implemented. This preliminary action evaluates data accuracy, identifies potential conflicts, and determines whether the confidence level meets the threshold required for safe change implementation, preventing harmful changes before they occur
Solution Approach 2:
The patent implements a feedback mechanism where confidence level results from data assessment feed back into the change decision process. This feedback loop provides continuous information about data reliability, enabling dynamic adjustment of change approval decisions and reducing the risk of adverse consequences from inaccurate information
3Reliability
If the system computes confidence levels and performs thorough verification before changes, then data reliability is improved, but the time required for change decision-making increases
Solution Approach 1:
The patent applies local quality assessment by computing confidence levels specifically for the configuration items affected by proposed changes, rather than verifying the entire CMDB. This localized approach focuses computational resources on relevant data, maintaining high reliability for decision-critical information while minimizing unnecessary verification time
Data Source
AI summary
In a configurable managed system having an associated configuration management database (CMDB), a Change Manager makes changes affecting configuration items (CIs), wherein different types of changes require different levels of confidence in the integrity of data in the CMDB. In response to a proposed change, weights assigned to system CIs affected by the change are used to compute a confidence level regarding accuracy of CMDB data pertinent to the change. The weight for a given CI is derived from the most recent synchronization thereof, and the numbers of reads, writes, and relationships the given CI has with other CIs. The confidence level is then used by the Change Manager in deciding whether or not to make the change. The proposed change is then allowed if the confidence level is no less than a pre-specified minimum value.