Virtualized Network Capacity Management via Entity Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtualized network environments, it is challenging to accurately monitor capacity usage and ensure efficient alignment between capacity supply and demand, leading to issues like stranded capacity and difficulties in fulfilling Service Level Agreements (SLAs), particularly due to the complexity of multi-layered virtualized networks and dynamic resource provisioning.
Innovation Solution
A method involving a capacity manager module that decomposes capacity entities into independent types, monitors availability, and adjusts resource allocation to align supply with demand, while also performing audits to detect and rectify variances and recurring allocation problems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If virtualization techniques are used to increase operational efficiency and dynamic resource provisioning, then resource utilization and energy efficiency are improved, but monitoring capacity usage and ensuring alignment between capacity supply and demand becomes increasingly difficult
Solution Approach 1:
The patent segments the complex virtualized network capacity into discrete capacity entities (CEs) that can be individually monitored and managed. Each CE represents a specific capacity unit that can be tracked independently, breaking down the overall complexity of monitoring virtualized resources into manageable segments.
Solution Approach 2:
The patent introduces capacity entity types (CETs) as intermediary abstractions between the physical infrastructure and the virtualized resources. These CETs serve as mediators that simplify monitoring by providing a standardized interface to track capacity usage across multiple virtualization layers without directly managing the underlying complexity.
2Adaptability or versatility
If multiple virtualized layers are implemented to accommodate dynamic workloads, then adaptability and service diversity are improved, but identifying and releasing stranded capacity becomes increasingly difficult
Solution Approach 1:
The patent implements feedback mechanisms that continuously monitor the state of capacity entities across virtualized layers and provide information about capacity usage and availability. This feedback loop enables the system to detect stranded capacity by comparing actual usage against allocated capacity and automatically trigger releases when capacity is no longer needed.
Solution Approach 2:
The patent enables the capacity management system to automatically identify and release stranded capacity without manual intervention. The system self-monitors capacity entity states, detects when capacity becomes stranded through the feedback mechanism, and automatically releases it back to the pool of available resources.
3Ease of operation
If conventional non-virtualized network solutions are used, then capacity monitoring is simpler, but they cannot be extended to accommodate virtualized networks employing multiple capacity layers
Solution Approach 1:
The patent creates a universal capacity entity framework that can operate across both conventional and virtualized network environments. The capacity entity types and monitoring mechanisms are designed to be agnostic to the underlying infrastructure, allowing the same system to manage capacity in traditional networks while also accommodating complex multi-layer virtualized networks.
Solution Approach 2:
The patent adds an abstraction dimension by introducing capacity entity types as a new layer of indirection between the physical infrastructure and the management system. This additional dimension allows the system to maintain the simplicity of conventional monitoring while extending capability to virtualized environments by operating at this intermediate abstraction level.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques are described for virtualized network capacity management. A hierarchical system includes a collection of capacity managers that can manage capacity providers providing resources such as compute resources, network resources, and/or storage resources to capacity consumers. The capacity managers can provide accurate capacity threshold reporting by decomposing any dependent capacity entities into a collection of independent capacity entities, thus enabling accurate available and unavailable capacity determinations. The capacity managers can implement capacity auditing procedures to detect variances between the intended and actual capacity allocation for each capacity provider. The capacity auditing can include audits between each capacity manager and the capacity providers that it manages, or audits between capacity managers. A cross-auditing procedure can be implemented to reconstruct capacity relationships from information in the capacity providers, and detect and isolate root causes of reoccurring problems that are usually only revealed when studying the results of multiple audits.