Coordinated Reliability Management for Virtual Machine Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional virtual machine placement solutions do not consider platform reliability and power budget attributes, leading to potential power budget violations and system downtimes, which are costly and can result in undesirable runtime actions.
Innovation Solution
A coordinated reliability management system that uses a cross-layer coordinated scheme to continuously match monitored platform parameters with virtual machine availability requirements, employing a multi-pass algorithm to select suitable host nodes that meet both virtual machine and platform metrics, and a coordinator agent to preemptively migrate virtual machines to more reliable hosts based on reliability models and service level agreements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional virtual machine placement solutions are used, then device complexity is reduced, but reliability deteriorates due to power budget violations and system downtimes
Solution Approach 1:
The system segments reliability management into separate modules: a reliability manager that monitors platform parameters, a coordinator that matches VM requirements with platform capabilities, and a virtualization manager that executes placements. This modular segmentation allows each component to focus on specific tasks, improving reliability while keeping individual component complexity manageable.
Solution Approach 2:
The coordinator acts as an intermediary between the reliability manager and the virtualization manager. It receives reliability data from the reliability manager, matches it with VM availability requirements, and translates into placement decisions for the virtualization manager. This intermediary layer coordinates information flow and decision-making across different system layers without requiring direct complex interactions between all components.
2Reliability
If virtual machines are placed without considering reliability metrics, then placement speed is improved, but reliability deteriorates leading to power budget violations
Solution Approach 1:
The reliability manager continuously monitors platform parameters (power consumption, temperature, reliability metrics) and maintains updated reliability models before VM placement decisions are made. The coordinator pre-matches VM availability requirements with predicted platform reliability, enabling proactive placement decisions that prevent power budget violations before they occur, rather than reacting after violations happen.
Solution Approach 2:
The system implements a feedback loop where the reliability manager continuously monitors actual platform parameters and compares them against predicted values from reliability models. This feedback information is fed back to the coordinator, which adjusts placement decisions accordingly. The feedback mechanism ensures power budget compliance by learning from actual system behavior and refining future placement predictions.
3Reliability
If continuous monitoring of platform parameters is implemented, then reliability is improved, but use of energy increases
Solution Approach 1:
The reliability manager monitors only the most critical platform parameters that directly impact reliability and power budget compliance, rather than continuously monitoring all possible system parameters. The system selectively monitors key metrics such as power consumption, temperature, and reliability indicators, reducing the energy cost of monitoring while maintaining adequate reliability awareness for effective VM placement decisions.
Data Source
AI summary
A system and methods for reliability management of virtual machines in a host system. The reliability of the host system is monitored and compared with a reliability threshold level for a virtual machine. If the reliability of the host system drops below the reliability threshold level, the virtual machine is migrated to another host system having an appropriate level of reliability.