Fabric-Backplane Server Provisioning via Virtualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current server technologies face challenges in dynamically provisioning and managing compute, storage, and network resources efficiently, leading to suboptimal datacenter availability, efficiency, and utilization.
Innovation Solution
The implementation of a hybrid server/multi-layer switch system architecture, known as the Enterprise Fabric (EF) architecture, which allows for dynamic provisioning of virtual servers with virtual networks, enabling flexible resource pooling, scaling, and reclamation, and supports high-bandwidth, low-latency I/O operations through virtualized compute, storage, and networking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional server provisioning methods are used, then hardware resources are allocated statically, but resource utilization efficiency deteriorates due to inability to dynamically adjust to changing demands
Solution Approach 1:
The patent implements dynamic resource provisioning by enabling virtual server instances to be created, modified, and destroyed on-demand based on workload requirements. The system allows runtime addition of compute, storage, and network resources to virtual servers without physical reconfiguration, transforming static hardware allocation into dynamic virtual resource management that adapts to changing demands in real-time
Solution Approach 2:
The patent creates a universal resource pool where physical hardware resources (compute nodes, storage arrays, network switches) can serve multiple virtual server instances simultaneously. The fabric-backplane architecture enables any virtual server to access any available resource through virtualization layers, allowing the same physical infrastructure to fulfill diverse computational, storage, and networking requirements across different applications and workloads
2Adaptability or versatility
If physical hardware reconfiguration is performed to adjust resources, then resource allocation can be modified, but system downtime and operational disruption increase
Solution Approach 1:
The patent creates virtual copies of physical hardware resources through virtualization. Virtual servers are instantiated as software-based representations that can be rapidly deployed, modified, and migrated without touching the underlying physical infrastructure. This copying approach allows resource reallocation to occur in the virtual layer while the physical hardware remains stable and operational, eliminating downtime associated with physical reconfiguration
Solution Approach 2:
The patent segments the monolithic physical server into multiple isolated virtual server instances running on the same hardware platform. Each virtual server operates as an independent unit with its own operating system and resource allocations, allowing individual instances to be modified, updated, or replaced without affecting other virtual servers or the host system, thereby maintaining continuous availability during resource adjustments
3Reliability
If excessive hardware resources are provisioned to handle peak demands, then service level requirements are met, but cost and resource waste increase during low-utilization periods
Solution Approach 1:
The patent merges multiple underutilized physical servers into a consolidated hardware platform running multiple virtual servers. By combining resources from several individual servers into a shared fabric-backplane infrastructure, the system achieves higher overall utilization rates while maintaining the ability to provision sufficient capacity for peak demands through virtual resource pooling and dynamic allocation across the merged infrastructure
Solution Approach 2:
The patent enables dynamic adjustment of resource allocation parameters for virtual servers based on actual workload conditions. Compute capacity, storage allocation, and network bandwidth can be modified in real-time through software configuration rather than fixed hardware settings, allowing the system to scale resources up during peak periods and scale down during low-utilization periods, thereby matching resource provision to actual service level requirements and eliminating energy waste from static over-provisioning
4Ease of operation
If manual provisioning and management processes are used, then resource allocation can be controlled, but provisioning time and operational complexity increase
Solution Approach 1:
The patent implements self-service automation where virtual servers can be provisioned, configured, and managed through software-based orchestration without manual hardware intervention. The system automatically allocates virtualized resources from the available pool, configures networking and storage connections, and manages lifecycle operations through automated workflows, eliminating the time-consuming manual processes of physical server assembly, cable routing, and configuration that previously required skilled technicians and extended provisioning cycles
Data Source
AI summary
Real time provisioning and management of fabric-backplane enterprise servers includes monitoring system status and configuration, displaying monitoring results, accepting user commands, and providing hardware and software management and configuration commands to the system. In one embodiment, an event is generated when a pluggable module is inserted into the system. In response to the event, the availability of the pluggable module is displayed to a system operator, and the operator enters a command to provision a server that includes the pluggable module. The server provisioning command is processed, resulting in a hardware configuration command being issued to the system, and an event indicating a status associated with processing the command is returned. The recognition of the inserted module, the display to the operator, and the processing of the server provisioning command occur in real time.