InfiniBand Target Groups for VM Migration Congestion Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high performance computing environments face challenges in managing live migration of virtual machines due to complex addressing and routing schemes in lossless networks like InfiniBand, leading to connectivity issues and performance degradation during VM migration.
Innovation Solution
Implement a system and method for congestion control in a private fabric using InfiniBand networks, which includes defining target groups with bandwidth limits, utilizing a target group repository, and employing dynamic reconfiguration mechanisms to minimize overhead and ensure seamless live migration of virtual machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If complex addressing and routing schemes are used in lossless networks like InfiniBand, then network performance is improved, but VM migration becomes difficult due to connectivity issues
Solution Approach 1:
The patent segments the network fabric into multiple virtual fabrics, allowing VM migration to occur within isolated virtual fabric segments. This segmentation enables independent management of migration traffic separate from production traffic, resolving the conflict between maintaining high network performance and enabling seamless VM migration.
Solution Approach 2:
The patent introduces a fabric manager as an intermediary component that handles the complexity of addressing and routing during VM migration. The fabric manager mediates between the VM migration process and the underlying InfiniBand network, abstracting the complex addressing schemes away from the migration operation while preserving network performance.
2Reliability
If bandwidth limits are defined on inter-switch links, then congestion control is improved, but resource allocation flexibility is reduced
Solution Approach 1:
The patent implements dynamic bandwidth allocation within virtual fabrics, where bandwidth limits are not fixed but can be adjusted based on current network conditions and migration requirements. This dynamic approach maintains reliable congestion control while preserving resource allocation flexibility, as the system can adapt bandwidth allocation in real-time during VM migration operations.
Solution Approach 2:
The virtual fabric infrastructure provides multi-functionality by serving both production workloads and migration traffic through the same physical infrastructure. The system universally applies congestion control mechanisms across different traffic types while maintaining the ability to allocate resources flexibly based on priority and current needs, resolving the contradiction between reliability and adaptability.
3Productivity
If live migration of virtual machines is implemented, then resource utilization is improved, but network overhead and downtime increase
Solution Approach 1:
The patent performs preliminary actions by pre-establishing virtual fabric connections and preparing migration paths before actual VM migration occurs. This preliminary setup within isolated virtual fabrics reduces the time required during actual migration, minimizing downtime while maintaining high resource utilization benefits.
Solution Approach 2:
The patent enables continuous operation during VM migration by maintaining active connections within the virtual fabric throughout the migration process. The migration occurs seamlessly within the isolated virtual fabric environment, allowing production workloads to continue without interruption while resource utilization is optimized through live migration capabilities.
Data Source
AI summary
Systems and methods for supporting target groups for congestion control in a private fabric in a high performance computing environment. An exemplary method can provide, at one or more microprocessors, a first subnet, the first subnet comprising a plurality of switches, a plurality of host channel adapters, and a plurality of end nodes, including a plurality of virtual machines. The method can define a target group on one of an inter-switch link or at a port of a switch of the plurality of switches, wherein the target group defines a bandwidth limit on the at least one of an inter-switch link between two switches of the plurality of switches or at a port of a switch of the plurality of switches. The method can provide a target group repository stored in a memory of the host channel adapter where the defined target group in the target group repository is recorded.