Fabric-Backplane Servers with Virtual Network Interface Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current server architectures face challenges in dynamically provisioning and managing compute, storage, and network resources to enhance datacenter availability, efficiency, and utilization, particularly in providing scalable and high-bandwidth, low-latency I/O operations while supporting multi-site fail-over and dynamic resource allocation.
Innovation Solution
The Enterprise Fabric (EF) architecture introduces a hybrid server/multi-layer switch system with Virtual Network Interface Controllers (vNICs), Virtual Local Area Networks (VLANs), and a switch fabric dataplane, enabling dynamic provisioning of virtual servers with virtual networks, scalable resource pooling, and high availability through modular components and redundant systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional server architectures are used, then hardware resources are dedicated and fixed, but resource utilization and adaptability are poor
Solution Approach 1:
The server system is segmented into multiple virtual servers, each with dedicated virtualized resources (CPU, memory, storage, I/O). This segmentation enables independent provisioning and management of resources for different applications while sharing physical hardware infrastructure, thereby improving adaptability without proportionally increasing overall system complexity.
Solution Approach 2:
The physical server hardware is designed to support multiple virtual server instances simultaneously, making the hardware universal and multi-functional. A single physical server can host multiple virtual servers with different configurations and requirements, improving resource utilization and adaptability while reducing the number of physical devices needed.
2Reliability
If physical servers are over-provisioned to handle peak loads, then availability is improved, but resource utilization efficiency decreases
Solution Approach 1:
Virtual server resources are dynamically allocable and reconfigurable. Resource allocation can be adjusted in real-time based on actual demand, allowing the system to maintain high availability during peak loads while efficiently utilizing resources during lower-demand periods. This dynamic provisioning eliminates the need for static over-provisioning.
Solution Approach 2:
The virtualization management system automatically provisions and manages resource allocation across multiple virtual servers without requiring manual intervention or over-provisioning. Resources are allocated on-demand to virtual servers that need them, ensuring availability when required while maintaining high overall utilization efficiency across the infrastructure.
3Adaptability or versatility
If dedicated hardware is allocated to each function, then performance is guaranteed, but scalability and dynamic reconfiguration are limited
Solution Approach 1:
A virtualization layer acts as an intermediary between physical hardware and virtual servers. This layer manages I/O resource allocation and can provide guaranteed performance through resource reservations and quality of service mechanisms while simultaneously enabling flexible dynamic reconfiguration of resources across virtual servers without dedicated hardware constraints.
Solution Approach 2:
The system allows dynamic changing of resource allocation parameters (CPU shares, memory allocation, storage I/O bandwidth, network bandwidth) for virtual servers without physical reconfiguration. These parameter changes enable adaptable resource provisioning while performance guarantees are maintained through virtual resource reservations and scheduling policies that ensure minimum bandwidth and processing guarantees.
Data Source
AI summary
Virtual Network Interface Controllers (vNICs) provide for communication among modules of Enterprise Server (ES) embodiments via a switch fabric dataplane. Processes executing on compute complexes of the servers exchange data as packets or messages by interfaces made available through vNICs. The vNICs further provide for transparent communication with network and storage interfaces. vNIC provisioning capabilities include programmable bandwidth, priority scheme selection, and detailed priority control (such as round-robin weights). In some embodiments, vNICs are implemented in Virtual Input/Output Controllers (VIOCs). In another aspect, Virtual Local Area Networks (VLANs) enable access to layer-2 and selected layer-3 network functions while exchanging the packets and messages. VLAN identification is provided in each vNIC, and VLAN processing is partially performed in VIOCs implementing vNICs. The compute complexes and interfaces are typically configured as pluggable modules inserted into a backplane included in a chassis.