VM-Aware Virtualization Architecture With Distributed Local Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional data center architectures are inflexible and costly, struggling to adapt to the dynamic nature of virtual machines in cloud environments, leading to performance bottlenecks and inefficiencies due to reliance on centralized storage management and networked storage solutions.
Innovation Solution
A scalable, VM-aware architecture that integrates solid-state drives and employs a distributed storage system with Controller VMs to manage storage resources, allowing local access and optimizing I/O operations, thus eliminating the need for network storage and enhancing performance and scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional network storage architecture is used, then storage resources can be shared across multiple servers, but performance deteriorates due to network bottlenecks and centralized storage management becomes a significant performance bottleneck as the number of storage devices increase
Solution Approach 1:
The patent segments the centralized storage management into distributed storage managers at each compute node. Each compute node has its own local storage devices and a corresponding storage manager that handles I/O operations locally, eliminating the network bottleneck associated with centralized storage management. This segmentation allows storage resources to remain shareable while performance is improved through local化管理.
Solution Approach 2:
The patent introduces local storage managers as intermediary components between compute nodes and storage devices. These local storage managers act as mediators that handle I/O operations locally rather than requiring all storage access to go through a centralized manager, thereby reducing network traffic and performance bottlenecks while maintaining resource sharing capabilities.
2Ease of operation
If traditional centralized storage management is used, then storage resources can be pooled and managed centrally, but device complexity increases and scalability is limited as the central storage manager becomes a significant performance bottleneck
Solution Approach 1:
The patent divides the centralized storage management architecture into multiple distributed storage managers, one at each compute node. This segmentation reduces the complexity burden on any single storage manager and enables better scalability. Each local storage manager manages only its own storage devices, simplifying the overall system architecture compared to a monolithic centralized manager.
Solution Approach 2:
The patent transitions from a single-point centralized storage management model to a distributed multi-point management model. By adding the dimension of distribution across multiple compute nodes, the system achieves better scalability and reduced complexity. The storage management function is replicated across multiple dimensions (nodes) rather than concentrated at a single point.
3Adaptability or versatility
If networked storage appliances are used, then storage can be accessed remotely, but cost increases due to expensive specialized rackmount or freestanding compute and storage devices
Solution Approach 1:
The patent merges storage devices with compute nodes, eliminating the need for separate specialized storage appliances. Storage resources are integrated directly into the compute nodes, allowing remote access through the virtualization layer. This merging approach reduces costs by using standard components rather than expensive specialized rackmount or freestanding storage devices.
Solution Approach 2:
The patent creates virtual copies of storage resources through virtualization. Instead of requiring physical remote storage access, the system creates virtual representations of storage that can be accessed by any compute node. This virtual copying approach eliminates the need for expensive specialized storage hardware while maintaining remote access capability.
4Stability of the object's composition
If physical machines are dedicated to single operating systems, then system stability is maintained, but resource utilization efficiency decreases during periods of inactivity
Solution Approach 1:
The patent introduces dynamic resource allocation through virtualization. Instead of static dedicated assignments, the system dynamically assigns physical resources to different virtual machines based on workload demands. The hypervisor manages these dynamic allocations, allowing operating systems to be migrated and resources to be reallocated in real-time, maintaining stability through controlled virtualization layers.
Solution Approach 2:
The patent creates universal physical hardware that can serve multiple operating systems simultaneously through virtualization. The physical machines become multi-functional platforms that can run different OS instances, replacing the traditional one-OS-per-physical-machine model. This universality maintains system stability through the hypervisor layer while dramatically improving resource utilization efficiency.
Data Source
AI summary
An improved architecture is provided which enables significant convergence of the components of a system to implement virtualization. The infrastructure is VM-aware, and permits scaled out converged storage provisioning to allow storage on a per-VM basis, while identifying I/O coming from each VM. The current approach can scale out from a few nodes to a large number of nodes. In addition, the inventive approach has ground-up integration with all types of storage, including solid-state drives. The architecture of the invention provides high availability against any type of failure, including disk or node failures. In addition, the invention provides high performance by making I/O access local, leveraging solid-state drives and employing a series of patent-pending performance optimizations.