Network-Aware Virtual Resource Provisioning Through Network Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud computing schedulers inefficiently provision virtual resources due to increased network traffic and resource utilization inefficiencies, leading to issues like packet drops and latency in data centers.
Innovation Solution
Collect and analyze network parameter data for network switches and physical servers to determine optimal deployment locations for virtual resources based on both hardware and network resource availability, using a network analysis platform to re-order rankings of physical servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If schedulers provision virtual resources based only on server hardware availability, then hardware resource utilization is optimized, but network performance deteriorates due to increased packet drops and latency
Solution Approach 1:
The patent segments the resource provisioning decision into two independent evaluation dimensions: hardware resource availability (CPU, memory, storage) and network resource availability (bandwidth, latency, packet drop rates). This segmentation allows the scheduler to independently optimize for each dimension and combine the results, resolving the contradiction between hardware utilization and network performance.
Solution Approach 2:
The patent applies local quality by evaluating network-specific parameters (bandwidth, latency, packet drop rates) at the network interface level rather than treating all servers uniformly. This allows the system to identify servers with superior network characteristics for specific workloads, thereby improving network performance without sacrificing hardware utilization efficiency.
2Productivity
If virtual resources are deployed to maximize hardware utilization, then computing efficiency improves, but network resource constraints are violated leading to packet drops
Solution Approach 1:
The patent performs preliminary evaluation of network resource availability and constraints before deploying virtual resources. By assessing bandwidth, latency, and packet drop rates in advance, the scheduler can pre-select servers that will not violate network resource constraints, thereby preventing packet drops before they occur while maintaining high computing efficiency.
Solution Approach 2:
The patent implements a feedback mechanism that continuously monitors network performance metrics (packet drop rates, latency) and uses this information to adjust future provisioning decisions. This feedback loop ensures that computing efficiency is maintained while network resource constraints are respected, as servers exhibiting harmful network behavior are automatically deprioritized in subsequent provisioning cycles.
3Device complexity
If traditional schedulers are used without network analytics, then system complexity remains low, but resource provisioning accuracy deteriorates
Solution Approach 1:
The patent introduces a network analysis platform as an intermediary component that bridges the gap between traditional schedulers and network resources. This intermediary collects and analyzes network performance data, then provides enhanced provisioning recommendations to the scheduler. This approach improves provisioning accuracy by incorporating network analytics while keeping the core scheduler relatively simple, as the complex analysis is performed by the dedicated intermediary platform.
Data Source
AI summary
This disclosure describes techniques for collecting network parameter data for network switches and/or physical servers and provisioning virtual resources of a service on physical servers based on network resource availability. The network parameter data may include network resource availability data, diagnostic constraint data, traffic flow data, etc. The techniques include determining network switches that have an availability of network resources to support a virtual resource on a connected physical server. A scheduler may deploy virtual machines to particular servers based on the network parameter data in lieu of, or in addition to, the server utilization data of the physical servers (e.g., CPU usage, memory usage, etc.). In this way, a virtual resource may be deployed to a physical server that has an availability of the server resources, but also is connected to a network switch with the availability of network resources to support the virtual resource.