Data Center Topology Mapping via NLP Process Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data center migration and management are complex due to the rapid evolution of data centers, requiring a detailed understanding of their topology across physical, power, cooling, cabling, and logical, virtualization, and software levels, which existing tools inadequately address, leading to laborious and error-prone migration processes.
Innovation Solution
A method for data center cartography generation using natural language processing (NLP) to identify computing nodes and their connections by refining rule sets, extracting known and unknown process entities from logs, and creating visual dependency representations, enabling a comprehensive and accurate mapping of data center systems and applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional data center mapping tools are used, then the process is simpler, but the accuracy and completeness of topology understanding deteriorates
Solution Approach 1:
The patent introduces process table data as an intermediary element that mediates between existing monitoring tools and the desired topology map. By extracting process entity information from process tables, the system achieves accurate topology mapping without requiring complex direct monitoring of all data center components, thus resolving the contradiction between accuracy and complexity
Solution Approach 2:
The patent replaces traditional mechanical/physical monitoring methods with information-based processing. Instead of physically tracing connections or using complex monitoring hardware, the system substitutes this with automated extraction and processing of process table data, achieving high-accuracy topology mapping through information analysis rather than physical measurement
2Productivity
If manual data collection methods are used, then the tool complexity is lower, but the time and labor required increases
Solution Approach 1:
The patent implements self-service by enabling the system to automatically extract and process topology information from process tables without human intervention. The NLP processing system autonomously identifies process entities, determines relationships, and generates the topology map, eliminating manual data collection efforts and significantly improving migration planning speed despite the added system complexity
Solution Approach 2:
The system incorporates feedback mechanisms where the NLP processor continuously refines its extraction rules based on the structure and content of process table data. This automated feedback loop allows the system to adapt and improve its topology mapping accuracy over time, maintaining high productivity while managing the complexity of the processing system
3Loss of information
If existing monitoring tools are used, then the implementation is easier, but the ability to capture detailed process relationships deteriorates
Solution Approach 1:
The patent applies extraction by specifically isolating and extracting process entity information from the broader process table data. By focusing only on the relevant process relationship information and excluding unnecessary details, the system captures comprehensive process relationships while managing data processing complexity through selective information extraction
Solution Approach 2:
The patent segments the complex process table data into distinct process entities and their relationships. By dividing the information into manageable components (processes, nodes, connections), the system can capture detailed process relationships without being overwhelmed by the overall complexity of the data structure
Data Source
AI summary
One embodiment provides a method including identifying all computing nodes and connections associated with the computing nodes in a data center based on running processes in the data center that communicate with one another. For each computing node, running processes are identified using natural language processing (NLP) by: iteratively refining a rule set that enables processing of surveillance information from the data center into an initial map of systems and applications in the data center, and extracting known process entities according to predetermined rules from the rule set. A visual dependency representation of the computing nodes and the processes running on the computing nodes is generated.