Abstract

It is important to assess availability of virtualized systems in IT business infrastructures. Previous work on availability modeling and analysis of the virtualized systems used a simplified configuration and assumption in which only one virtual machine (VM) runs on a virtual machine monitor (VMM) hosted on a physical server. In this paper, we show a comprehensive availability model using stochastic reward nets (SRN). The model takes into account (i) the detailed failures and recovery behaviors of multiple VMs, (ii) various other failure modes and corresponding recovery behaviors (e.g., hardware faults, failure and recovery due to Mandelbugs and aging-related bugs), and (iii) dependency between different subcomponents (e.g., between physical host failure and VMM, etc.) in a virtualized servers system. We also show numerical analysis on steady state availability, downtime in hours per year, transaction loss, and sensitivity analysis. This model provides a new finding on how to increase system availability by combining both software rejuvenations at VM and VMM in a wise manner.

Highlights

  • Computing systems with virtualization are rapidly gaining strong attention for computational sustainability by administrators of information resources in enterprises

  • In order to study system characteristics in terms of business availability and continuity featured for computational sustainability in an IT business infrastructure, we analyzed the following metrics: steady-state availability (SSA), transaction loss, and sensitivity of the steady state availability (SSA) with respect to clocks’ interval

  • We have modeled and analyzed a virtualized servers system with multiple virtual machine (VM) via stochastic reward nets (SRN)

Read more

Summary

Introduction

Computing systems with virtualization are rapidly gaining strong attention for computational sustainability by administrators of information resources in enterprises. It is necessary to incorporate various hardware and software failure modes along with corresponding recovery behaviors and analyze the availability of such systems. The main drawbacks of previous work are that most of virtualized systems are composed of only one VM running on one VMM in a physical server; see the papers [9, 10]. This architecture is commonly used in modeling and analysis of a virtualized server system in a number of studies, even though the proposed architecture in their hypothesis shows multiple VMs hosted on one VMM [11, 12].

Related Work
A Virtualized Server System
SRN Models of the VSS
PVM2clock
Numerical Results and Discussions
PVMMclock
VI VII
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call