Network Manager Allocating Virtual Machines in Multi-Tenant Datacenters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-tenant datacenters face performance variability issues due to shared resources, leading to unpredictable application performance and increased costs for tenants and providers, which makes them unsuitable for applications requiring predictable performance and reduces datacenter efficiency and revenue.
Innovation Solution
Implementing a network manager that allocates virtual machines based on available slots and performance requirements, using virtual network abstractions like virtual clusters and virtual oversubscribed clusters to provide guaranteed bandwidth and efficient resource utilization, while updating residual capacity values for subsequent requests and enforcing performance characteristics through rate-limiting mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If resources are shared between multiple tenants in a datacenter, then resource utilization efficiency is improved, but performance variability increases making it unsuitable for applications requiring predictable performance
Solution Approach 1:
The patent segments the physical datacenter network into multiple virtual networks, each dedicated to a specific tenant. This segmentation isolates tenant traffic so that one tenant's activities cannot affect another's performance, eliminating performance variability while maintaining high resource utilization through virtualization. Each virtual network gets guaranteed bandwidth and performance characteristics.
Solution Approach 2:
The patent introduces a network manager as an intermediary between tenants and the physical datacenter infrastructure. This network manager allocates resources, enforces performance guarantees, and manages virtual network configurations. It acts as a mediator that ensures each tenant receives their guaranteed performance while optimizing overall resource utilization across the shared infrastructure.
2Reliability
If performance guarantees are provided to tenants through dedicated network resources, then performance predictability is improved, but resource utilization efficiency and datacenter revenue decrease
Solution Approach 1:
The patent creates virtual networks that can be universally applied to multiple tenants with different performance requirements. Each virtual network template can be replicated and configured for different tenants, providing standardized performance guarantees while allowing the underlying physical infrastructure to be shared and dynamically allocated across multiple users, maintaining both predictability and efficiency.
Solution Approach 2:
The patent enables dynamic adjustment of network parameters such as bandwidth allocation, latency requirements, and packet loss thresholds for different virtual networks. This allows the system to provide performance guarantees tailored to specific tenant needs while optimizing resource utilization by allocating parameters efficiently across the shared infrastructure based on actual demand patterns.
3Speed
If virtual machines are allocated without considering performance requirements, then allocation speed is improved, but network performance variability increases
Solution Approach 1:
The patent performs preliminary actions by pre-configuring virtual network templates with defined performance characteristics and bandwidth allocations before tenants request resources. When tenants request virtual machines, the system can quickly allocate them to appropriate pre-configured virtual networks that already have performance guarantees established, enabling fast allocation while maintaining performance consistency through the pre-established virtual network configurations.
Data Source
AI summary
Methods of offering network performance guarantees in multi-tenant datacenters are described. In an embodiment, a request for resources received at a datacenter from a tenant comprises a number of virtual machines and a performance requirement, such as a bandwidth requirement, specified by the tenant. A network manager within the datacenter maps the request onto the datacenter topology and allocates virtual machines within the datacenter based on the available slots for virtual machines within the topology and such that the performance requirement is satisfied. Following allocation, stored residual capacity values for elements within the topology are updated according to the new allocation and this updated stored data is used in mapping subsequent requests onto the datacenter. The allocated virtual machines form part of a virtual network within the datacenter which is allocated in response to the request and two virtual network abstractions are described: virtual clusters and virtual oversubscribed clusters.