Service Function Chaining via Time-Aware Resource Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional distributed routing protocols in multi-domain networks fail to consider resource availability, leading to unpredictable delays and high computation complexity in service function chaining, especially in cloud and big data applications where resources are limited and geographically distributed.
Innovation Solution
A vertex-centric distributed approach that uses a resource orchestration framework to identify qualified service function chaining solutions by determining available vertices, mapping service functions, and provisioning resources based on time availability, ensuring efficient resource allocation and reduced delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional distributed routing protocols are used to compute network paths, then routing simplicity is maintained, but resource availability is not considered leading to unpredictable delays
Solution Approach 1:
The patent introduces a resource orchestrator as an intermediary component that sits between the distributed routing protocols and the service function chaining infrastructure. This orchestrator maintains a global view of resource availability and coordinates resource allocation across multiple domains, enabling predictable delays without requiring complex modifications to the underlying distributed routing protocols themselves
Solution Approach 2:
The system segments the resource management functionality into distributed vertex agents at each network node and a centralized resource orchestrator. Each vertex agent handles local resource status reporting and receives allocation decisions, while the orchestrator makes global decisions based on aggregated information, dividing the complexity management across multiple levels
2Productivity
If a centralized orchestrator maintains a hierarchical topology for all domains, then service function chaining can be coordinated, but computation complexity increases significantly on a large-scale global view
Solution Approach 1:
The patent divides the global network into multiple domains, each with its own hierarchical topology maintained by domain-specific orchestrators. This segmentation allows the global computation complexity to be distributed across multiple smaller domain-level computations, reducing the burden on any single centralized orchestrator while maintaining coordinated service function chaining across domains
Solution Approach 2:
Domain orchestrators act as intermediaries between the global resource orchestrator and local network infrastructure. They aggregate resource information from their domains and present it to the global orchestrator, while receiving and enforcing global allocation decisions, thereby reducing the computational burden on the global view through hierarchical abstraction
3Adaptability or versatility
If virtual machines are used to migrate network functions to multi-purpose platforms, then service flexibility improves, but setup times become unpredictable due to queuing delays
Solution Approach 1:
The resource orchestrator performs preliminary actions by pre-provisioning and pre-configuring virtual machine templates and resource allocations based on predicted service function chaining requirements. By preparing resources in advance and maintaining a pool of pre-configured virtual machine images, the system reduces actual setup times when service functions need to be deployed, making them more predictable while preserving the flexibility of virtualization
Solution Approach 2:
The system implements feedback mechanisms where vertex agents continuously report resource availability and usage status to the resource orchestrator. This real-time feedback enables the orchestrator to dynamically adjust virtual machine provisioning and resource allocation, predicting and preventing queuing delays by reallocating resources before delays occur, thus making setup times more predictable while maintaining service flexibility
Data Source
AI summary
A resource orchestration framework may include vertices representing physical network nodes and a resource orchestrator. The resource orchestrator may receive a service function chain request specifying service functions to perform on respective physical nodes and a first starting time for launching a first one of the service functions. The resource orchestrator may determine an ending time for performing the first service function that, with the first starting time, defines a time duration for the first service function, identify vertices at which the first service function is available during the time duration, map a first vertex to the first service function in a candidate service function chain, determine a second starting time for launching a second one of the service functions dependent on the first starting time, and provide the candidate service function chain, the first starting time, and the second starting time to a neighbor vertex of the first vertex.