Virtual Machine Live Migration Memory Transfer Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current live migration technologies face challenges in efficiently estimating the memory dirty rate and transfer time during virtual machine migration, leading to potential service outages and increased operational overhead, which can exceed acceptable limits and impact service level agreements.

Innovation Solution

A method utilizing access counters to estimate the memory transfer time for live migration by counting memory accesses and calculating the dirty rate, allowing for precise planning and execution of virtual machine migrations with minimal overhead on the virtual machine's operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional live migration methods are used, then virtual machine migration can be performed, but service outage time may exceed acceptable limits due to inability to accurately estimate memory transfer time

Engineering Contradiction:
Improveservice level agreement complianceVSAvoidservice outage time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by estimating the memory dirty rate and transfer time before initiating live migration. The management program calculates these parameters in advance using monitoring data from the virtual machine's memory access patterns, allowing the system to predict whether migration will complete within acceptable service outage limits before actually performing the migration operation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback by continuously monitoring memory access patterns and dirty page generation rates during virtual machine operation. This real-time feedback data is used to dynamically adjust migration timing decisions and predict transfer completion times, enabling the system to make informed decisions about when to initiate migration to minimize service impact.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If memory dirty rate monitoring is performed using conventional methods, then transfer time estimation becomes possible, but operational overhead increases significantly

Engineering Contradiction:
Improvememory transfer time estimation accuracyVSAvoidoperational overhead
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies self-service by utilizing the virtual machine's own memory management structures and existing page fault handling mechanisms to gather dirty rate information. The management program leverages already-present system components such as the page table and memory management unit, avoiding the need for separate monitoring hardware or complex external measurement systems.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent implements universality by using the same memory management infrastructure that serves the virtual machine's operational needs also for migration estimation purposes. The page fault counters and memory access tracking that are inherently part of the virtual memory system are repurposed to provide dirty rate measurement, eliminating the need for dedicated monitoring mechanisms.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP2804100B1Virtual machine management program, method and device
Publication Date: 2020.04.29 FUJITSU LTD
  • EP2804100B1 patent drawingFigure 1
  • EP2804100B1 patent drawingFigure 2
  • EP2804100B1 patent drawingFigure 3

AI summary

Information useful for a live migration can be obtained during usual operations of a virtual machine to make it possible to perform efficient planning and execution of the live migration. A non-transitory computer-readable recording medium has a program stored therein for causing a computer to execute a process of managing the live migration of the virtual machine. The process includes estimating a cost of executing the live migration of the virtual machine, using a count value of an access counter for counting the number of accesses to a memory allocated to the virtual machine, a capacity of the memory, and a bandwidth of data transfer between physical machines relating to the live migration.