Workload Placement Using Logical Network Resource Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing workload placement techniques in SDN environments often fail to consider logical network resource utilization, leading to network failures due to exceeded limits on logical switches, routers, and other resources, despite focusing on physical resource utilization.
Innovation Solution
A resource optimization system collects and processes logical network resource capacity and utilization information to determine the realization cost of placing a virtual computing instance (VCI) on a host, selecting a target host that does not exceed logical network resource limits, and migrates the VCI accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If workload placement focuses on physical resource utilization (processor and memory), then physical resource efficiency is improved, but logical network resource limits are exceeded causing network failures
Solution Approach 1:
The patent extends the workload placement parameters beyond traditional physical resources (CPU, memory) to include logical network resources (logical switches, routers, flow table entries). The system calculates a composite score that incorporates both physical resource utilization metrics and logical network resource consumption metrics, thereby changing the parameter set used for decision-making in workload placement.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring logical network resource utilization on each host and using this information to adjust workload placement decisions. The logical network resource utilization data is fed back into the scoring algorithm, allowing the system to dynamically adapt placement decisions based on current network resource states, preventing exceedance of logical resource limits.
2Adaptability or versatility
If more logical network resources are allocated to hosts to support additional VCIs, then workload placement flexibility is improved, but logical network resource utilization increases risking limit exceedance
Solution Approach 1:
The system performs preliminary assessment of logical network resource requirements before making workload placement decisions. By calculating the realization cost (the amount of logical network resources that would be consumed) for each potential host in advance, the system can predictively determine whether placing a workload on a particular host would exceed logical resource limits, thereby preventing network failures before they occur.
Solution Approach 2:
The patent introduces new parameters for measuring and tracking logical network resource utilization (number of logical switches, routers, flow table entries) and integrates these into the workload placement scoring mechanism. This parameter expansion allows the system to balance workload placement flexibility against logical network resource consumption, optimizing both aspects simultaneously.
Data Source
AI summary
The disclosure provides an approach for workload migration. Embodiments include receiving logical network resource capacity information and logical network resource utilization information relating to a plurality of host computers and to one or more logical network resources. Embodiments include determining that a virtual computing instance (VCI) is to be run on one of the plurality of host computers and determining for each respective host computer of the plurality of host computers, a respective realization cost of the VCI for the respective host computer, wherein the respective realization cost relates to the one or more logical network resources. Embodiments include selecting, based on the logical network resource capacity information, the logical network resource utilization information, and the realization cost, a target host computer for the VCI from the plurality of host computers and loading the VCI on the target host computer.