Stateless Virtual Machine Traffic Shaping via Packet Drop Probability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cloud computing platforms face high CPU overhead due to the need for maintaining queues and state information to manage shared resources, which can lead to resource starvation and inefficiencies in bandwidth allocation among virtual machines.
Innovation Solution
A stateless mechanism is implemented in the cloud computing platform to shape virtual machine communication traffic by monitoring average communication rates, calculating drop probabilities, and randomly dropping communication messages based on these probabilities, thereby reducing overhead and ensuring equitable resource access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional procedures (leaky bucket, token bucket, fair queuing) are used to share resources equitably, then resource fairness is improved, but CPU overhead increases significantly due to maintaining queues and state information
Solution Approach 1:
The patent extracts and removes the queue and state information maintenance from the traffic shaping mechanism. Instead of maintaining complex queues and state for each virtual machine, the system uses a simple counter that is incremented for each packet and a threshold-based dropping mechanism, eliminating the computational overhead while preserving fairness
Solution Approach 2:
The patent replaces expensive, long-lived queue structures with cheap, short-lived counter variables. The counter is simply incremented and compared against a threshold, and then reset or adjusted, creating a lightweight stateless mechanism that requires minimal CPU resources while achieving the same traffic shaping objectives
2Reliability
If state information is maintained for traffic shaping, then equitable resource access is improved, but computational overhead increases
Solution Approach 1:
The patent implements a self-service mechanism where the counter automatically increments with each packet and naturally provides fairness through its simple arithmetic operation. The system does not require complex state management or external control logic - the counter itself performs the traffic shaping function through straightforward increment and threshold comparison operations
Solution Approach 2:
The patent changes the state representation from complex queue structures to a simple numerical counter parameter. This parameter is easily manipulated through increment operations and threshold comparisons, reducing computational complexity while maintaining the ability to enforce equitable resource access through the threshold-based dropping mechanism
3Ease of operation
If queues are maintained for traffic shaping, then resource allocation control is improved, but system efficiency decreases due to resource starvation risks
Solution Approach 1:
The patent inverts the traditional approach by not maintaining queues at all, but instead using a counter that tracks packet flow and triggers dropping when thresholds are exceeded. This inverted approach simplifies the system while maintaining control, and actually improves efficiency by eliminating the queue management overhead and reducing the risk of resource starvation through stateless operation
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Cloud computing platforms having computer-readable media that perform methods to shape virtual machine communication traffic. The cloud computing platform includes virtual machines and a controller. The controller limits the traffic associated with the virtual machines to enable the virtual machines to achieve desired communication rates, especially when a network servicing the virtual machines is congested. The controller may drop communication messages associated with the virtual machines based on a drop probability evaluated for the virtual machines.