NFV Virtual Machine Instance Co-Location for Service Chain Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In existing network architectures, multiple service processing devices are connected in series, leading to increased processing capacity demands, unnecessary network device investments, and fault points, resulting in inefficient service flow processing and potential interruptions when a service processing device fails.
Innovation Solution
Deploying virtual machine instances of value-added servers (VAS) and service switches (SSW) on the same physical machine, with predefined communication relationships, to reduce communication requirements and enhance processing efficiency by minimizing inter-machine communication and optimizing service chain processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple service processing devices are connected in series, then service flow processing can be performed, but processing capacity demand increases and fault points increase
Solution Approach 1:
The patent segments service flows into different service chains based on user subscriptions and service types. Each service chain is independently routed through specific service processing devices, so that not all service flows pass through all devices. This segmentation reduces the processing capacity demand on each individual device and eliminates single points of failure, as a fault in one service chain does not affect other service chains.
2Adaptability or versatility
If multiple service processing devices are connected in series, then comprehensive service processing is achieved, but unnecessary network device investment increases
Solution Approach 1:
The patent implements a universal service chain routing mechanism that can dynamically direct different service flows through different sequences of service processing devices based on user subscriptions and service types. This multi-functional routing approach allows the same physical network infrastructure to support diverse service processing requirements without needing dedicated devices for each service type, thereby reducing unnecessary network device investment while maintaining comprehensive service processing capability.
3Productivity
If VAS instances and SSW instances are deployed on different physical machines, then resource distribution is achieved, but communication capability requirements increase
Solution Approach 1:
The patent combines VAS instances and SSW instances that have communication relationships onto the same physical machine. This merging eliminates the need for inter-machine communication between these components, thereby reducing communication capability requirements and network overhead. The consolidation improves service chain processing efficiency by allowing direct local communication between VAS and SSW instances while maintaining resource distribution across multiple physical machines for other components.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for deploying a virtual machine instance is disclosed, so as to lower requirements for a communication capability of a virtualized value-added server and improve processing efficiency of a service chain. A corresponding apparatus and device are further provided. The method includes: obtaining communication relationships between a value-added server VAS instances and a service switch SSW instances from a service template, where the VAS instances and the SSW instances are used to provide services in a service chain (301), and the service chain and the communication relationships between the VAS instances and the SSW instances are defined in the service template; and deploying, according to the communication relationships, an SSW instance and a VAS instance that need to communicate with each other in the SSW instances and the VAS instances on a same physical machine (302).