VM Migration for Network Congestion Relief
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for managing network link utilization in data centers are complex and time-consuming, requiring significant network topology changes and signaling to optimize traffic flow, especially when migrating Virtual Machines (VMs) to alleviate congestion, which can lead to prolonged service outages and increased costs.
Innovation Solution
A system and method that uses a modeling function to monitor traffic matrices and network topology to determine the minimum number of VMs to migrate and their optimal locations, allowing for intelligent redistribution of traffic without altering the network topology, thereby optimizing network utilization and reducing migration time and costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional methods are used to manage network link utilization by migrating VMs, then network congestion can be alleviated, but the process becomes complex and time-consuming requiring significant network topology changes and signaling
Solution Approach 1:
The patent extracts the traffic engineering functionality from the network topology itself and places it in the VM placement layer. Instead of modifying network topology to route traffic around congestion, the system migrates VMs to different physical locations so that traffic naturally flows through less congested links, eliminating the need for complex network topology changes and signaling mechanisms
Solution Approach 2:
Conventional approaches invert the problem by trying to control traffic flow through network routing mechanisms. The patent inverts this by controlling VM placement decisions, which indirectly controls traffic patterns. By placing VMs strategically based on traffic matrix analysis, the system achieves traffic engineering without direct network routing manipulation
2Productivity
If VM migration is performed to optimize network traffic distribution, then bandwidth utilization improves, but migration time and service outage increase
Solution Approach 1:
The system performs preliminary analysis of the traffic matrix and identifies optimal VM placement decisions before actual migration occurs. By pre-calculating the best migration scenarios and their impacts on network utilization, the system can execute migrations more efficiently with minimal service disruption, as the optimal paths and timing are already determined
Solution Approach 2:
The patent applies partial migration strategies where only the minimum necessary VMs are migrated to achieve the desired network optimization. By carefully selecting which VMs to move and to where, the system achieves improved bandwidth utilization without the excessive time cost of migrating all VMs or performing comprehensive network reconfiguration
3Adaptability or versatility
If network topology is engineered to support both normal routing behavior and optimized traffic flow, then traffic engineering capability improves, but network complexity and costs increase
Solution Approach 1:
The patent extracts traffic engineering capability from the network layer and implements it at the VM placement layer. The system uses a simplified network topology that maintains normal routing behavior while achieving traffic optimization through intelligent VM placement decisions based on traffic matrix analysis, eliminating the need for complex network engineering
4Adaptability or versatility
If MPLS and virtual links are used to direct traffic around congested links, then reactive traffic engineering is enabled, but network complexity and signaling requirements increase
Solution Approach 1:
The patent extracts reactive traffic engineering capability from MPLS and network signaling mechanisms and implements it through VM placement decisions. By analyzing the traffic matrix and strategically placing VMs on different physical hosts, the system achieves reactive traffic engineering without requiring MPLS labels, virtual links, or complex network signaling protocols
Solution Approach 2:
The patent replaces the mechanical network signaling and routing mechanisms (MPLS, virtual links) with a software-based VM placement optimization approach. The traffic engineering is achieved through computational optimization of VM placement decisions rather than through network protocol mechanisms, significantly reducing network complexity
Data Source
AI summary
A system and method that facilitates the migration of one or more Virtual Machines (VMs) throughout a communications network, such as a cloud network, is disclosed. A management system monitors a current traffic matrix to determine if the network is congested. If the network is congested, the management system determines a minimum number of VMs that must be migrated to unused VM sites to ease the congestion and optimize the distribution of traffic in the network. Additionally, the management system also identifies which VMs should be migrated to which unused VM sites, and a sequence in which the migration should be performed to return the bandwidth distribution in the network to an acceptable level. In addition, the management system may determine which unused VM site or sites is best for the addition of one or more new VMs or groups of VMs.