Network Availability Scoring for Co-location Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Co-location facilities face challenges in evaluating and ensuring high availability of network connection services across multiple layers of the network stack, leading to potential service interruptions and failures that can breach Service-Level Agreements (SLAs).
Innovation Solution
A system is developed to assess high availability by applying metrics to network and application infrastructure across multiple layers, providing individualized scoring for each feature and determining a fine-grained rating for network connection services, which can be displayed to operators to aid in infrastructure upgrades or reconfigurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If network infrastructure components and features are frequently added to meet customer demands, then service functionality and adaptability are improved, but system complexity and difficulty in evaluating high availability increase
Solution Approach 1:
The patent segments the complex network infrastructure evaluation into multiple OSI layers (physical, data link, network, transport, session, presentation, application). Each layer is evaluated independently with layer-specific metrics, allowing the system to handle complexity through modular assessment rather than treating the entire infrastructure as a monolithic system.
Solution Approach 2:
The patent introduces a multi-dimensional evaluation framework that assesses high availability across seven OSI layers simultaneously. This dimensional approach transforms the complex single-problem evaluation into a structured multi-layered assessment, where each dimension (layer) contributes to the overall high availability score through weighted aggregation.
2Measurement precision
If comprehensive high availability evaluation across multiple OSI layers is performed, then reliability assessment accuracy is improved, but computational resources and evaluation time are consumed
Solution Approach 1:
The patent implements selective layer evaluation where not all seven OSI layers need to be fully assessed in every evaluation cycle. The system can focus on critical layers based on service type and risk assessment, performing comprehensive evaluation only when necessary. This partial action approach maintains accuracy for critical assessments while reducing routine evaluation overhead.
Solution Approach 2:
The system performs preliminary assessments using simplified metrics before conducting full multi-layer evaluations. Quick health checks and baseline measurements are taken continuously, allowing the system to skip detailed evaluation for stable infrastructure segments and focus computational resources only on areas requiring detailed assessment.
3Loss of information
If individualized scoring for each network device feature is implemented, then identification of specific improvement areas is improved, but data processing complexity and computational overhead increase
Solution Approach 1:
The patent assigns different scoring weights and evaluation criteria to different OSI layers and network features based on their specific importance to high availability. Critical features like routing redundancy and failover mechanisms receive higher weights, while less critical features receive lower weights. This local quality approach allows detailed identification of improvement areas without uniformly processing all features at maximum complexity.
Solution Approach 2:
The system dynamically adjusts evaluation parameters such as scoring weights, threshold values, and metric priorities based on service type, infrastructure criticality, and historical performance data. This parameter adaptation allows the same evaluation framework to handle diverse network configurations efficiently, reducing computational overhead by adjusting the granularity of assessment to match actual needs.
Data Source
AI summary
Techniques are disclosed for evaluating and facilitating high availability for network connection services provided by network infrastructure of one or more co-location facilities. For example, a computing system receives data from each network device in one or more co-location facilities, the data indicating the respective capability of each network feature of each network device to provide highly available operation across a plurality of communication model layers for supporting a network connection service to customer networks co-located within the one or more co-location facilities. The computing system applies high availability evaluation metrics to the data to determine a high availability capability score for each network device feature of each network device. Further, the computing system determines, based on the high availability capability scores, an indication of the high availability for the network connection service. The computing system outputs, for display, the indication of the high availability of the network connection service.