An Novel Approach to Evaluate the Reliability of Cloud Rendering
System Using Probabilistic Model Checker PRISM : A Quantitative
Computing Perspective
Haoyu Liu
1,3
, Huahu Xu
1,3*
1
School of Computer Engineering and Science, Shanghai
University, 200444 Shanghai, China, liuhaoyu76@163.com
2
Computing Center, Shanghai University, 200444 Shanghai,
China, gaohonghao@shu.edu.cn
Honghao Gao
1,2
, Minjie Bian
3
, Huaikou Miao
1,4
3
Shanghai Shang Da Hai Run Information System Co. Ltd,
200444 Shanghai, China, huahuxu@163.com
4
Shanghai Key Laboratory of Computer Software Testing &
Evaluating, 201112 Shanghai, China, hkmiao@shu.edu.cn
ABSTRACT
This paper proposes an approach to evaluate the reliability of
cloud rendering system. After the requirement analysis, the
rendering system was divided into three modules: preparing files,
requesting resources, and rendering task execution. Each module
may have an exception that will reduce reliability, and has the
ability to recover it. To expose these details, the discrete-time
Markov chain (DTMC) is improved to formalize the cloud
rendering system. The model contains an abnormal state set
representing exceptions and errors such as file corruption and
failure to rendering subtasks. Then, a series of formal properties
are defined to describe reliability in detail. The proposed method
gives full consideration to the processes of rendering tasks.
Finally, the properties are verified by performing PRISM in a
quantitative way. The experiment shows that our method is
effective to evaluate the reliability of the cloud rendering system.
CCS Concepts
•Software and its engineering → Software verification;
Software reliability;
Keywords
cloud rendering; DTMC model checking; PRISM; reliability
1. INTRODUCTION
Rendering is the relatively important step in the production of
film and television cartoons. Cloud rendering [20,21,23] is a
popular method used by filmmakers to perform rendering tasks
to reduce time and save money. To ensure quality of service
(QoS) [24,25], the service provider usually proavides a service
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or
distributed for profit or commercial advantage and that copies bear this notice and
the full citation on the first page. Copyrights for components of this work owned
by others than ACM must be honored. Abstracting with credit is permitted. To
copy otherwise, or republish, to post on servers or to redistribute to lists, requires
prior specific permission and/or a fee. Request permissions from
permissions@acm.org.
MobiQuitous 2017, November 7–10, 2017, Melbourne, VIC, Australia
© 2017 Association for Computing Machinery.
ACM ISBN 978-1-4503-5368-7/17/11$15.00
https://doi.org/10.1145/3144457.3144460
level agreement to clarify the QoS parameters and
responsibilities to be carried out. There are various definitions of
QoS attributes [1,2,9]. Software reliability is the most
commonly mentioned attribute. However, it is not perceivable
by the client. Service providers need to test their system and
verify reliability by themselves.
There are two ways to evaluate software reliability: the
black-box method and the white-box method [23]. The black-
box method uses a mathematical statistics method but does not
consider the system internal structure and operational processes.
It makes the reliability definition general, and the evaluation is
not detailed. The cloud rendering system is complex, and it is
not sufficient to define and verify reliability from only a holistic
perspective. However, the white-box approach is based on the
path and state of the system model and uses model checking for
a quantitative analysis. It takes full account of the details.
In this paper, we use a white-box approach, and improved a
model to evaluate the reliability of cloud rendering systems.
Compared to other models, this model is useful for presenting
more detailed reliability attributes. Based on the approach, the
reliability checking results will become more specific and more
useful for analyzing the reliability of cloud rendering systems.
Our work is divided into the following points. First, an extended
discrete-time Markov chain (DTMC) model is proposed. In the
proposed model, the states are divided into two categories:
normal and abnormal. The set of abnormal states contains some
error states describing the cause of failure. Then, nine reliability
properties are defined from three aspects, including success rate,
exception rate and recovery rate. Where the success rate aspect
analyzes mainly the running results of the entire system, the
other two are aimed at analyzing specific function modules.
Finally, these properties are verified by using PRISM [6,7]. The
experimental results are used to evaluate the reliability of the
three aforementioned aspects. Experiments show that the
proposed method can describe the reliability of a cloud
rendering system in more detail and guarantee the validity of the
reliability check.
The rest of this paper is structured as follows: Section 2
describes the related works. Section 3 proposes an extended
DTMC to model the cloud rendering system. Section 4 defines a
set of reliability properties and corresponding formulas in
PRISM. Section 5 verifies the effectiveness of the proposed
method by experiment. Section 6 summarizes the research
contents of this paper and suggests a future research direction.