This paper develops an availability model of a virtualized system. We construct non-virtualized and two hosts system models using two-level hierarchical approach in which fault trees are used the upper level homogeneous continuous time Markov chains (CTMC) to represent sub-models lower level. In models, we incorporate not only hardware failures (e.g., CPU, memory, power, etc) but also software including Virtual Machine Monitor (VMM), (VM), application failures. high (HA) service VM live...

Consolidated server systems using virtualization involves serious risks of host failures that induce unexpected downs all hosted virtual machines and applications. To protect applications requiring high-availability from unpredictable failures, redundant configuration can be an effective countermeasure. This paper presents a machine placement method for establishing against with less servers. The proposed estimates the requisite minimum number according to performance requirements...

As server virtualization is used as an essential software infrastructure of various services such cloud computing, availability management virtualized system becoming more significant. Although time-based rejuvenation useful to postpone/prevent failures due aging in a system, the schedules for virtual machine (VM) and monitor (VMM) need be determined proper way VM availability, since VMM affects VMs running on VMM. This paper presents analytic models using stochastic reward nets three...

Server virtualization is a technology used in many enterprise systems to reduce operation and acquisition costs, increase the availability of their critical services. Virtualized may be even more complex than traditional nonvirtualized systems; thus, quantitative assessment system difficult. In this paper, we propose sensitivity analysis approach find parameters that deserve attention for improving systems. Our based on Markov reward models, suggests host failure rate most important...

High-availability assurance of cloud service is a critical and challenging issue for providers. To quantify the availability services from both architectural operational points views, modeling evaluation are essential. This paper presents component-based framework, named Candy, which constructs comprehensive model semi-automatically system specifications described by Systems Modeling Language (SysML). SysML diagrams translated into components assembled together to form entire in Stochastic...

The Multi-access Edge Computing (MEC) and Network Function Virtualization (NFV) integrated architecture is a key enabling platform for 5G to run multiple customized services in the form of service function chain (SFC) configured as an ordered set functions (SFs). However, memory-related software aging SF that can be exploited by attackers becomes new threat dependability MEC-SFC services. To provide dependable services, proactive rejuvenation techniques counteract problem are essential. In...

This article analyzes the completion time of a job running on virtualized server subject to software aging and rejuvenation in virtual machine monitor (VMM). A may be interrupted by (VM) failure, VMM failure or rejuvenation. The interruption is categorized as either preemptive-repeat ( prt ), which case needs restart from beginning, preemptive-resume prs resumes execution point interruption. Using semi-Markov process (SMP) model behavior, steady-state availability computed theory developed...

Software aging is a phenomenon of progressive degradation software execution environment caused by faults. In this paper, we propose life-extension as an operational countermeasure against and present the mathematical foundations means stochastic modeling. A semi-Markov process used to capture behavior system with analyze system's availability completion times jobs running on it. The can correctly model time-based allows us derive optimal trigger for starting in terms mean job time. We also...

Cloud computing is established on advanced software technologies intended to enhance the scalability of infrastructure by making full use commodity servers. The more cloud relies technologies, bugs have significant impacts system availability. Aging-related bugs, which cause accumulation errors in long time execution, often remain even after release a stable version software. This paper investigates bug reports five major open-source projects related and confirms issues about existence...

Network Function Virtualization (NFV) enables network operators to utilize a common substrate for offering different networking services flexibly and dynamically in the form of service chains. A chain NFV consists set Virtualized Functions (VNFs) that are interconnected through logical service. VNFs usually implemented on top virtual machines hosted commodity hardware with reliabilities. Ensuring end-to-end reliability is crucial providing highly-available In this paper, we propose embedding...

Software aging (i.e. progressive performance degradation of long-running software systems) is difficult to detect due the long latency until it manifests during program execution. Fast and accurate detection important for eliminating underlying defects already development testing. Also in a deployment scenario, needed plan mitigation methods like rejuvenation. The goal this paper evaluate whether Mann-Kendall test an effective approach detecting from traces computer system metrics. This...

Image classifiers for recognizing real-world objects are widely used in the Internet of Things (IoT) and Cyber-Physical Systems(CPSs). A classifier is trained offline by machine learning algorithms with training data sets, then it deployed on a cloud or an edge computing system online label predictions. As classifier's performance depends underlying software infrastructure, may degrade over time due to faults causing aging. In this paper, we address issue experimentally investigate aging...

Software systems running continuously for a long time often confront software aging, which is the phenomenon of progressive degradation execution environment caused by latent faults. Removal such faults in development process crucial issue system reliability. A known major obstacle typically large latency to discover existence aging. We propose systematic approach detect aging has shorter test and higher accuracy compared traditional detection via stress testing trend with high confidence....

Virtualization technology has promoted the fast development and deployment of cloud computing, is now becoming an enabler Internet Everything. Virtual machine monitor (VMM), playing a critical role in virtualized system, software hence it suffers from aging after long continuous running as well crashes due to elusive faults. Software rejuvenation techniques can be adopted reduce impact aging. Although there existed analytical model-based approaches for evaluating techniques, none analyzed...

This paper addresses the coverage (including identification and isolation) of irrelevant components in systems with imperfect fault (IFC). In fault-tolerant systems, a single not-covered component may thwart automatic recovery mechanisms, lead to system or subsystem failure. The models that consider effects IFC are known as (CMs). traditional CMs, except those considering functional dependency (a similar concept relevancy but different assumptions semantics), is typically limited faulty...

In modern IT systems, data backup and restore operations are essential for providing protection against loss from both natural man-made incidents. On the other hand, can be resource-intensive lead to performance degradation, or may require system offline entirely. Therefore, it is important properly choose techniques policies ensure adequate while minimizing impact on availability performance. this paper, we present an analytical modeling approach such a purpose. We study file service that...

Virtualization enables data centers to consolidate servers improve resource utilization and power consumption. This paper presents the issues of performability management in a virtualized center that hosts multiple services using virtualization. One main cause degradation is software aging both virtual machines (VMs) machine monitor (VMM) Virtualized Data Center (VDC). Software rejuvenation cost effective proactive method counteract aging. During operations, there occurs down time,...

This paper presents a high-availability solution to virtualized data center which consists of cluster physical servers and hosted virtual machines (VMs). VMs are typically generated through software called machine monitor (VMM) running on servers. Both VMMs face the risks aging caused by aging-related bugs in may result failures long time execution. To prevent such proactively, we propose combined server rejuvenation technique that performs VM rejuvenations simultaneously with VMM...

This paper presents a performability model for RAID storage systems using Markov regenerative process to compare different architectures. While homogeneous models are extensively used reliability analysis of systems, the memory-less property sojourn time assumed in such is not satisfied reality, especially disk rebuild whose progress interrupted even at an event another failure. In this paper, we use which allows us generally distributed times providing needed extension traditional models....

<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</i> ulti-access xmlns:xlink="http://www.w3.org/1999/xlink">e</i> dge xmlns:xlink="http://www.w3.org/1999/xlink">c</i> omputing (MEC)-enabled xmlns:xlink="http://www.w3.org/1999/xlink">I</i> nternet xmlns:xlink="http://www.w3.org/1999/xlink">o</i> f xmlns:xlink="http://www.w3.org/1999/xlink">T</i> hings (IoT) is considered as a promising paradigm to deliver computation-intensive and...

Virtual machine monitor (VMM) rejuvenation is a proactive recovery method against failures caused by software aging in VMM. Since the job running on hosted virtual (VM) interrupted at VMM rejuvenation, preemption type of important for performance execution. In preemptive repeat (prt), needs to restart after recovery. resume (prs), can execution from point. this paper, we analyze completion time virtualized server subject and prt case or prs case. The theory developed [1] used obtain...