SlideShare a Scribd company logo
A Novel Robust & Fault tolerance Framework for
Webservices using WS-I
*
Specification
Akhilesh Kumar Pandey
1,3
, Abhishek Kumar
1,4
and Farahnaz Rezaeian Zade
2,5
1
School of Computer Engineering , KIIT University, Bhubaneswar, Orissa, India
2
Fooland Institute Of Technology, Fooland Shahr, Isfahan, Iran
infoakhileshpandey@gmail.com
3
, abhikumar695@gmail.com
4
, fr.rezaeeian@gmail.com
5
Abstract— Fault handling never became trivial task for the
applications since decades as it is abnormal situation that may
occur when application running in a production environment. So
coming to the distributed application component such as
webservice also fault handling needs more sophisticated
approach to ensure QoS of service. Recently, Webservices has
become a defacto standard for developing loosely coupled
distributed application component due to the support of
interoperability irrespective of the platform and language
however still webservice architecture fails to provide robustness
and fault tolerance so that business organizations can provide
services to the consumer more accurately and effectively without
failing 24*7 hours. This work proposes a robust and fault
tolerance framework to handle faults in webservices and repair
the faults in webservices in order to provide smooth and efficient
services. A set of diversity and repair actions are also proposed
that provides normal resolution of services from manual
breakdown of services. In the framework all the stack of
webservice standards are used along with fault tolerance
capability in existing webservice architecture. Mainly, proposed
framework uses two base component that is Agent for
maintaining replicas and Controller for fault detection, fault
notification and fault confinement in order to provide robustness
and fault tolerance in existing webservice architecture.
Keywords—SOA; Robust; Fault Handling; Webservices;
SOAP; Interoperability; Fault tolerance; SCA.
I. INTRODUCTION
SOA(Service-Oriented Architecture) is an architectural
paradigm that improves accuracy, correctness & efficiency of
business organizations in order to fulfill business requirement
[1]. Service is a self contained interoperable loosely coupled
distributed component. There is a set of technologies to
implement Service Oriented Architecture (SOA) but
webservices is the best way to realize SOA. Web service
technology follows some stack of standards such as WSDL,
SOAP, which uses XML to provide interoperability [2-4].
SOA also can be implemented through SCA (Service
Component Architecture) [5].
Usually, Service oriented systems are made by
heterogeneous webservices that are distributed across network
and provided by business organizations . Developing highly
reliable, actuate and efficient service oriented system is tedious
task due to the dynamic behavior of webservices. Webservices is
the technology that allows us to build interoperable, loosely
coupled, distributed software component, It also allows a
application to expose complex business logic over the network.
Webservice software component are not only platform
independent, these are even language independent which means it
allows to develop software component using Perl, Python, C,
C++, Java or .NET . In order to access this consumer component
can be developed in any one of technology like C, C++, Java or
.NET.
To build interoperable distributed component, every
software vender started defining their own stack of standards
such as transport protocol, language for exchanging data and
description. In such cases interoperability will never become
reality. Later, it was realized by every software vendor present
in the market that the fact that they can't build interoperable
component.
Currently, Webservice have become very important part of
critical application due to this availability of application
becomes dependent on the webservices used by the
applications. QoS & availability is a very important part of
critical application. However, there is no stack of standards for
robustness and fault tolerance in webservices. Considering
robustness and fault tolerance as an critical issue , there many
attempt have been taken place by the industry and academic
researchers but still their proposal lacks to provide enough
support for robustness and fault tolerance as their work don't
take the advantage of specification given by WS-I [6] .
The proposed work focuses on the robustness and fault
tolerance(FT) in webservices. The main idea in this paper is to
propose framework that can provide high QoS (quality of
service) and availability. Proposed framework takes an
advantage of existing webservice architecture and WS-I
specification in order to ensure compatibility with stack of
webservice standards[6]. The proposed framework takes an
advantage of webservice architecture along with two new
major components that is Controller and Agent.
The rest of paper is organized as follows , related work on
fault tolerance in distributed technologies and Webservices is
described in section 2. A basic idea about the technology is
presented in section 3. A novel robust and fault tolerance
framework using WS-I specification is proposed in section 4,
it also includes the description about two major components of
proposed framework. Description of proposed framework with
step by step presented in section 5. In section 6 conclusion and
future work have been described.
II. RELATED WORK
With the help of related data and information, it is a
universally accepted fact that fault tolerance can be achieved
with duplicating hardware and software resources [7] [8]. For
instance, if services are duplicated and one of them fails, the
remaining one can still ensure the desired output.
Due to the high demand of web services, issues related to
Fault Tolerance are becoming critical. Therefore an
appropriate solution is needed to solve such type of issues. In
the recent years, several fault handling methodologies and
frameworks have been proposed for web services and SOA.
But due to a high demand of web services it still requires a
more theatrical background and robust and fault tolerant
framework to provide high availably and quality of service .
Vijay Dialani et al.[9] have proposed a transparent fault
tolerance for Web services based architectures which provide
capacity for monitoring the individual instance and service
host. Chen Liu et al.[10] have proposed a rule-based exception
handling approaches in SOA, in which they introduced some
basic rules, categories of exception with strategies to handle it.
Huang et al.[11] have proposed a runtime monitoring
composite web service through stateful aspect extension where
they followed stateful aspect extension approach to monitor
web service at runtime. Roman Vaculfn et al.[12] have
proposed exception handling model with appropriate recovery
actions for semantic web services that is purely based on
OWL-S. Wen jiajia et al.[13] have proposed a system for
handling exception in BPEL process through multi-policy, but
proposed work lacks to provide reliability and extensibility of
web services. M.G. In [14] author proposed the mechanism of
recovering faulty Web apps using Service discovery in which
they described the mechanism for handling faults using
various kinds of the recovery mechanism. In their work they
introduced a self healing architecture for identifying faults and
taking appropriate actions based on faults. In [15], service
oriented architecture(SOA) based service recovery framework
has been proposed. In this work's author introduced an
algorithm to select recovery action based on exception
servility but it still needs various recovery actions and
compatibility based on web service stack of standards such as
WSDL, SOAP and UDDI. M. Panahi et al.[16] have defined
the a middleware architecture to support an accountable
service oriented architecture that consists of three components
ASB, AA and Agent. Their work mainly focuses to mange
web service faults that to only those are closely related to
performance. But in their work various kinds of faults in
services are not considered. B. Pernici et al.[17] proposed a
work for identifying the fault type and methodologies to fix
various types of faults. They categorized the various types of
faults based on state such as permanent, transient, which are
determined by Bayesian classification. However, their work
requires more concrete emphasis to identify the type of fault
related to fault duration and criteria to choose methodologies.
Tamir et al. [18] proposed a fault tolerance model for
webserver which detects server errors and forwards requests to
a replica server to avoid service failures. In [19], Author
proposed a fault tolerant architecture for grid applications.
This model takes a help of a passive replication approach in
order to provide fault tolerance.
III. TECHNOLOGY BACKGROUND
This section contains basic concepts of web services
related to the fault handling in distributed technologies and
web services.
Initially We have applications which are designed to run on
one machine. Like when we develop applications in 'C' language ,
it can execute on a box and can talk to other applications which
are running on the same box. But now distributed technology
solves the above problems and allows to build applications which
is not only executed on a machine and talk to other applications
on the same machine, these applications should also be able to
access or communicate with other applications which are running
on a different machine.
The web service is technology that allows us to build an
interoperable distributed application component. To build
interoperable distributed application component, every
software vendor in the industry started defining their own
standards. In such case interoperability never will become
reality. With a time it was realized by every software vendor
that we can't build independently an interoperable distributed
application component, unless everyone has agreed upon same
standards. There comes a WS-I[6] organization which is
formed to build open standards for building interoperable
distributed application component. WS-I released BP1.0
specification document, the document contains a set of
guidelines or standards that must be used to develop a
distributed interoperable component .
Web services have three main parts in its architecture [1-
3,5] that is a service provider, service consumer and UDDI
registry [21].In order to achieve interoperability in
webservices common language for exchange data is used that
is XML but having XML is not enough to exchange data
between two systems. We need a layer on top of it to classify
the information. For instance, We don't send only business
data for communication, along with business data, we may
need to send transport specific data or some helper data. So
using XML we can't differentiate such kind of information.
Hence in such cases SOAP is used that can differentiate
business data and helper data . In the proposed framework we
use a SOAP binding protocol for interaction.
This paper focuses in the field of robustness and FT(Fault
Tolerance) in webservices. The key idea is to provide a
framework that can offer QoS with high availability of web
services. Proposed framework uses the existing webservice
architecture to ensure compatibility with stack of webservices
standards.
IV. PROPOSED FRAMEWORK
Our framework is based on the existing web service
architecture and WS-I organization specification with fault
tolerance and robust capability in order to provide QoS and
achieve interoperability up to the mark.
A. Avizienis et al.[20] given basic concept and dependable
and secure computing in which different types of faults are
discussed that may arise during service invocation.
The proposed framework includes two major components
that are responsible to mange replica for diversity actions,
fault diagnosis and appropriate recovery actions.
In our proposed framework, configuration stage is
responsible for publication of web services where the provider
of service uses the UDDI registry for registering web service
replica.
On execution stage, an Agent module is used to create
web-service replicas. Once the web - service replica is created
Agent performs different test to ensure the service that has to
be used. In this stage Controller job is to detect the faults. It
also does the verification of the web service during its
execution. If a fault gets propagated, the Controller
communicates with Agent for selection of appropriate web
service that is replicated across various nodes.
There are two major modules of proposed framework that is
Controller and Agent .
1. Controller: Controller module in framework
provides QoS to consumers by taking different kind
of necessary actions. It also gives facility to provide
for detecting faults and taking actions based upon
faults.
Figure 1: A Novel Robust & Fault tolerance Framework for Webservices using WS-I* Specification
The main responsibility of Controller for fault
detection, notification and confinement. It gives
notification to the Agent if webservice renders with a
fault during the service execution. For example, if
webservice gets a fault during its execution, then
Controller gives fault notification to the Agent in
order to assign appropriate web service replica to
continue normal execution from manual breakdown.
In the framework Controller gets a request from
consumer with an appropriate web service type,
including operations and parameters for service
execution. Then Controller takes an help of Agent in
order to perform mapping of service type in
available webservice replica.
2. Agent: The main job of the agent is to manage replicas
across the nodes. It creates service replica using
UDDI registry [21]. The Agent also checks presence
of replica during replica selection stage.
V. DESCRIPTION OF PROPOSED FRAMEWORK
In the figure 1.0, There can be a separate Controller for
each and every web service consumer. In diagram Agent and
UDDI can be located remotely at different locations. Here
multiple Agent can be deployed in different locations with the
Controller. The proposed framework work as presented in
following steps:-
1. During execution of application, the consumer sends
a request of required type.
2. Consumer request passes to Agent through
Controller.
3. Agent searches web-service of required service type
from UDDI registry [21].
4. Now Agent performs an availability test for
choosing the appropriate web service replica that has
to be used.
5. Next, Agent returns web-service address of the
controller. Now the consumer can invoke the needed
operation/ method in the selected service through the
returned access point.
6. The Controller does the verification of service
execution. During execution, if any, faults occur,
then controller takes an help of replica and continues
the manual execution of the application.
7. Now, when service execution is completed results
are reported back to the consumer.
Here, the replica creation takes an advantage of UDDI
standards [21]. The extended UDDI allows a refined web-
service replica selection that is based on the availability,
interoperability and response time .
VI. CONCLUSION AND FUTURE WORK
In this paper a robust and fault tolerant framework for web
services using WS-I
*
specifications is proposed. Our
framework uses existing web service architecture, including
two more major component Controller and Agent to ensure
robustness and fault tolerance in web services. The main
benefit of web-service technology is interoperability
irrespective of the language and platform through open stack
of web service standards such as SOAP, WSDL, XML etc.
In future work, the idea is to propose a load balancing
algorithm for avoiding from consumer to forward multiple
requests to the same web service replica. Further work is also
to propose fault tolerance framework for mobile web services
in which proposed work may use REST instead of SOAP as
smart devices will have very less resources in terms of
memory. So by keeping hardware as well as software
resources as an issue, the idea is there to use a REST
interaction style of web services for mobile devices instead of
traditional SOAP.
ACKNOWLEDGMENT
This work is provisionally supported by KIIT University,
Bhubaneswar, India. Our sincere thanks to KIIT University,
Bhubaneswar, India for providing resources to complete this
work.
REFERENCES
[1] E. Thomas. SOA Principles of Service Design. Prentice Hall
International, 1st edition, 2007.
[2] W3C. Simple Object Access Protocol (SOAP)
1.1. http://www.w3.org/TR/soap/.
[3] W3C. Web Services Description Language (WSDL).
http://www.w3.org/TR/wsdl.
[4] OASIS. Universal Description, Discovery and Integration(UDDI).
http://uddi.org/.
[5] B. Margolis and J. Sharpe. SOA for the Business Developer - Concepts,
BPEL, and SCA. MC Press, 1st edition, 2007.
[6] WS Interoperability Organization. Web Service Interoperability (WS-I),
2005.
[7] B. Kim, “Reliability analysis of real-time controllers with dual-modular
temporal redundancy,” Proc. of the Sixth International Conference on
Real-Time Computing Systems and Applications (RTCSA)1999, pp.364-
371, 13-15 Dec. 1999.
[8] B. Kim, “Reliability analysis of real-time controllers with dual-modular
temporal redundancy,” Proceeding of the 6th International Conference
on Real-Time Computing Systems and Applications (RTCSA) 1999,
pp.364-371, 13-15 Dec. 1999.
[9] V. Dialani, S. Miles, L. Moreau, D. D. Roure, and M. Luck.
"Transparent fault tolerance for Web services based architectures". In
Proceedings of the 8th International Euro-Par Conference on Parallel
Processing, pages 889–898. Springer, 2002.
[10] Chen Liu, Yanting Xu, D., and Xaiyu, L. “A rule- based exception
handlings approach in SOA”. International Conference on Computer
Application and System Modeling (2010).
[11] Huang T, Wu GQ, Wei J, "Runtime Monitoring Composite Web
Services Through Stateful Aspect Extension", Journal Of Computer
Science And Technology, vol. 24, pp. 294- 308, 2009.
[12] Roman Vaculfn, Kevin Wiesner, Katia Sycara "exception handling and
recovery of semantic web services", 4th International Conference on
Networking and Services, ICNS 2008, Pages 217-222 ,2008.
[13] Wen Jiajia, Chen Junliang, Peng Yong and Xu Meng. "a multi policy
exception handling system for bpel process", First Intl. Conf. on
Communications and Networking in China, ChinaCom '06, 2007.
[14] M.G. Fungini and E. Mussi, “recovery of faulty web applications
through service discovery,” Proc. First International Workshop Semantic
Matchmaking and Resource Retrieval: Issues and Perspectives (SMR
2006), Sept. 2006, pages. 67-80.
[15] C. Jie-Ying, W. Yong-Jun, and X. Yi, “soa-based service recovery
framework,” Proceeding the 9th International Conference. Web-Age
Information Management (WAIM 2008), July 2008, pages. 629-635.
[16] M. Panahi, K.J. Lin, Y. Zhang, S.H. Chang, J. Zhang, and L. Varela,
“The LLAMA Middleware Support for Accountable Service-Oriented
Architecture,” Proceeding the 6th International Conference Service
Oriented Computing (ICSOC 2008), Lecture Notes in Computer Science
5364, Springer- Verlag, Dec. 2008, pp. 180-194.
[17] B. Pernici, and A.M. Rosati., “Automatic Learning of Repair Strategies
for Web Services,” Proceeding. 5th IEEE European Conf. Web services
(ECOWS 2007), IEEE Computer Society, Nov. 2007. Pages. 119-128.
[18] N. Aghdaie and Y. Tamir. "Client transparent fault-tolerant web
service". In Proceedings of the IEEE International Conference on
Performance, Computing, and Communications, pages 209-216. IEEE
Computer Society, 2001.
[19] D. Zagorodnov, X. Zhang, M. Hiltunen, K. Marzullo and R. D.
"Schlichting. Fault-tolerant grid services using primary backup:
feasibility and performance". In Proceedding IEEE International
Conference. on Cluster Computing, pages 105-114. IEEE-CS, 2004.
[20] A. Avizienis, J. C. Laprie, B. Randell, C. Landwehr. Basic concepts
and taxonomy of dependable and secure computing. IEEE Transactions
on Dependable and Secure Computing, pages 11–33, 2004.
[21] UDDI specifications,
http://uddi.org/pubs/uddiv3.0.2-20041019.html

More Related Content

PDF
AA using WS vanZyl 2002-05-06
DOC
Surya_CV
PDF
15 falko menge--_enterpise_service_bus
PDF
IRJET - Application Development Approach to Transform Traditional Web Applica...
PDF
Performance in soa context
PPT
WebServices and Workflow technologies
PDF
MuCon 2015 - Microservices in Integration Architecture
PPSX
ESB Overview
AA using WS vanZyl 2002-05-06
Surya_CV
15 falko menge--_enterpise_service_bus
IRJET - Application Development Approach to Transform Traditional Web Applica...
Performance in soa context
WebServices and Workflow technologies
MuCon 2015 - Microservices in Integration Architecture
ESB Overview

What's hot (19)

PDF
7 eai-patterns
PPT
EAI (Integration) and Mulesoft
PDF
Research Inventy : International Journal of Engineering and Science
PPSX
Getting started with Enterprise Application Integration (EAI) using Enterpris...
PPTX
Twelve factor-app
PDF
Finite State Machine Based Evaluation Model For Web Service Reliability Analysis
PDF
2011-ESB-WP-Draft
PDF
QOS OF WEB SERVICE: SURVEY ON PERFORMANCE AND SCALABILITY
PDF
Enterprise Application Integration
PPTX
Uunit 5-xml&web security
PPTX
Unit 3-SOA Technologies
PDF
Location-Aware and Personalized Collaborative Filtering for Web Service Recom...
PDF
WEB SERVICE COMPOSITION IN DYNAMIC ENVIRONMENT: A COMPARATIVE STUDY
PDF
Placement of BPM runtime components in an SOA environment
PPT
ESB Concepts
PDF
fidelity_cs_final
DOCX
PRASANTHI P
PPTX
Introduction to Enterprise Service Bus
PDF
6 ijmecs v7-n1-5 a novel testing model for soa based services
7 eai-patterns
EAI (Integration) and Mulesoft
Research Inventy : International Journal of Engineering and Science
Getting started with Enterprise Application Integration (EAI) using Enterpris...
Twelve factor-app
Finite State Machine Based Evaluation Model For Web Service Reliability Analysis
2011-ESB-WP-Draft
QOS OF WEB SERVICE: SURVEY ON PERFORMANCE AND SCALABILITY
Enterprise Application Integration
Uunit 5-xml&web security
Unit 3-SOA Technologies
Location-Aware and Personalized Collaborative Filtering for Web Service Recom...
WEB SERVICE COMPOSITION IN DYNAMIC ENVIRONMENT: A COMPARATIVE STUDY
Placement of BPM runtime components in an SOA environment
ESB Concepts
fidelity_cs_final
PRASANTHI P
Introduction to Enterprise Service Bus
6 ijmecs v7-n1-5 a novel testing model for soa based services
Ad

Similar to A Novel Robust &Fault Tolerance Framework for Webservices using ws-I Specification (20)

PDF
A Novel Framework for Reliable and Fault Tolerant Web Services
PDF
Enhancement in Web Service Architecture
PDF
Evaluation of a Framework for Integrated Web Services
PDF
Testing of web services Based on Ontology Management Service
PDF
Reliability evaluation model for composite web services
PDF
Framework for service oriented development of monolithic legacy software
PDF
PDF
fault tolerance management in cloud computing
PDF
A NOVEL METHOD TO TEST DEPENDABLE COMPOSED SERVICE COMPONENTS
PDF
Performance Evaluation of Web Services In Linux On Multicore
PDF
20507-38933-1-PB.pdf
PDF
COMPOSITE DESIGN PATTERN FOR FEATUREORIENTED SERVICE INJECTION AND COMPOSITIO...
PDF
Composite Design Pattern For Feature- Oriented Service Injection And Composit...
PDF
Contract Versioning
DOCX
Service oriented software engineering
PDF
Study on Use Case Model for Service Oriented Architecture Development
PDF
DOCX
Study on Use Case Model for Service Oriented Architecture Development
PDF
QOS OF WEB SERVICE: SURVEY ON PERFORMANCE AND SCALABILITY
PDF
AN ADAPTIVE APPROACH FOR DYNAMIC RECOVERY DECISIONS IN WEB SERVICE COMPOSITIO...
A Novel Framework for Reliable and Fault Tolerant Web Services
Enhancement in Web Service Architecture
Evaluation of a Framework for Integrated Web Services
Testing of web services Based on Ontology Management Service
Reliability evaluation model for composite web services
Framework for service oriented development of monolithic legacy software
fault tolerance management in cloud computing
A NOVEL METHOD TO TEST DEPENDABLE COMPOSED SERVICE COMPONENTS
Performance Evaluation of Web Services In Linux On Multicore
20507-38933-1-PB.pdf
COMPOSITE DESIGN PATTERN FOR FEATUREORIENTED SERVICE INJECTION AND COMPOSITIO...
Composite Design Pattern For Feature- Oriented Service Injection And Composit...
Contract Versioning
Service oriented software engineering
Study on Use Case Model for Service Oriented Architecture Development
Study on Use Case Model for Service Oriented Architecture Development
QOS OF WEB SERVICE: SURVEY ON PERFORMANCE AND SCALABILITY
AN ADAPTIVE APPROACH FOR DYNAMIC RECOVERY DECISIONS IN WEB SERVICE COMPOSITIO...
Ad

Recently uploaded (20)

PPTX
522797556-Unit-2-Temperature-measurement-1-1.pptx
PDF
Decoding a Decade: 10 Years of Applied CTI Discipline
PPTX
Funds Management Learning Material for Beg
PPTX
Power Point - Lesson 3_2.pptx grad school presentation
PDF
Cloud-Scale Log Monitoring _ Datadog.pdf
PPTX
INTERNET------BASICS-------UPDATED PPT PRESENTATION
PDF
The Internet -By the Numbers, Sri Lanka Edition
PDF
Slides PDF The World Game (s) Eco Economic Epochs.pdf
PPTX
June-4-Sermon-Powerpoint.pptx USE THIS FOR YOUR MOTIVATION
PDF
APNIC Update, presented at PHNOG 2025 by Shane Hermoso
PPTX
CHE NAA, , b,mn,mblblblbljb jb jlb ,j , ,C PPT.pptx
PDF
Automated vs Manual WooCommerce to Shopify Migration_ Pros & Cons.pdf
PPTX
Slides PPTX World Game (s) Eco Economic Epochs.pptx
PPTX
introduction about ICD -10 & ICD-11 ppt.pptx
PPTX
E -tech empowerment technologies PowerPoint
PDF
Tenda Login Guide: Access Your Router in 5 Easy Steps
PPTX
Digital Literacy And Online Safety on internet
PPTX
Introuction about ICD -10 and ICD-11 PPT.pptx
PDF
Vigrab.top – Online Tool for Downloading and Converting Social Media Videos a...
PPTX
Introduction about ICD -10 and ICD11 on 5.8.25.pptx
522797556-Unit-2-Temperature-measurement-1-1.pptx
Decoding a Decade: 10 Years of Applied CTI Discipline
Funds Management Learning Material for Beg
Power Point - Lesson 3_2.pptx grad school presentation
Cloud-Scale Log Monitoring _ Datadog.pdf
INTERNET------BASICS-------UPDATED PPT PRESENTATION
The Internet -By the Numbers, Sri Lanka Edition
Slides PDF The World Game (s) Eco Economic Epochs.pdf
June-4-Sermon-Powerpoint.pptx USE THIS FOR YOUR MOTIVATION
APNIC Update, presented at PHNOG 2025 by Shane Hermoso
CHE NAA, , b,mn,mblblblbljb jb jlb ,j , ,C PPT.pptx
Automated vs Manual WooCommerce to Shopify Migration_ Pros & Cons.pdf
Slides PPTX World Game (s) Eco Economic Epochs.pptx
introduction about ICD -10 & ICD-11 ppt.pptx
E -tech empowerment technologies PowerPoint
Tenda Login Guide: Access Your Router in 5 Easy Steps
Digital Literacy And Online Safety on internet
Introuction about ICD -10 and ICD-11 PPT.pptx
Vigrab.top – Online Tool for Downloading and Converting Social Media Videos a...
Introduction about ICD -10 and ICD11 on 5.8.25.pptx

A Novel Robust &Fault Tolerance Framework for Webservices using ws-I Specification

  • 1. A Novel Robust & Fault tolerance Framework for Webservices using WS-I * Specification Akhilesh Kumar Pandey 1,3 , Abhishek Kumar 1,4 and Farahnaz Rezaeian Zade 2,5 1 School of Computer Engineering , KIIT University, Bhubaneswar, Orissa, India 2 Fooland Institute Of Technology, Fooland Shahr, Isfahan, Iran infoakhileshpandey@gmail.com 3 , abhikumar695@gmail.com 4 , fr.rezaeeian@gmail.com 5 Abstract— Fault handling never became trivial task for the applications since decades as it is abnormal situation that may occur when application running in a production environment. So coming to the distributed application component such as webservice also fault handling needs more sophisticated approach to ensure QoS of service. Recently, Webservices has become a defacto standard for developing loosely coupled distributed application component due to the support of interoperability irrespective of the platform and language however still webservice architecture fails to provide robustness and fault tolerance so that business organizations can provide services to the consumer more accurately and effectively without failing 24*7 hours. This work proposes a robust and fault tolerance framework to handle faults in webservices and repair the faults in webservices in order to provide smooth and efficient services. A set of diversity and repair actions are also proposed that provides normal resolution of services from manual breakdown of services. In the framework all the stack of webservice standards are used along with fault tolerance capability in existing webservice architecture. Mainly, proposed framework uses two base component that is Agent for maintaining replicas and Controller for fault detection, fault notification and fault confinement in order to provide robustness and fault tolerance in existing webservice architecture. Keywords—SOA; Robust; Fault Handling; Webservices; SOAP; Interoperability; Fault tolerance; SCA. I. INTRODUCTION SOA(Service-Oriented Architecture) is an architectural paradigm that improves accuracy, correctness & efficiency of business organizations in order to fulfill business requirement [1]. Service is a self contained interoperable loosely coupled distributed component. There is a set of technologies to implement Service Oriented Architecture (SOA) but webservices is the best way to realize SOA. Web service technology follows some stack of standards such as WSDL, SOAP, which uses XML to provide interoperability [2-4]. SOA also can be implemented through SCA (Service Component Architecture) [5]. Usually, Service oriented systems are made by heterogeneous webservices that are distributed across network and provided by business organizations . Developing highly reliable, actuate and efficient service oriented system is tedious task due to the dynamic behavior of webservices. Webservices is the technology that allows us to build interoperable, loosely coupled, distributed software component, It also allows a application to expose complex business logic over the network. Webservice software component are not only platform independent, these are even language independent which means it allows to develop software component using Perl, Python, C, C++, Java or .NET . In order to access this consumer component can be developed in any one of technology like C, C++, Java or .NET. To build interoperable distributed component, every software vender started defining their own stack of standards such as transport protocol, language for exchanging data and description. In such cases interoperability will never become reality. Later, it was realized by every software vendor present in the market that the fact that they can't build interoperable component. Currently, Webservice have become very important part of critical application due to this availability of application becomes dependent on the webservices used by the applications. QoS & availability is a very important part of critical application. However, there is no stack of standards for robustness and fault tolerance in webservices. Considering robustness and fault tolerance as an critical issue , there many attempt have been taken place by the industry and academic researchers but still their proposal lacks to provide enough support for robustness and fault tolerance as their work don't take the advantage of specification given by WS-I [6] . The proposed work focuses on the robustness and fault tolerance(FT) in webservices. The main idea in this paper is to propose framework that can provide high QoS (quality of service) and availability. Proposed framework takes an advantage of existing webservice architecture and WS-I specification in order to ensure compatibility with stack of webservice standards[6]. The proposed framework takes an advantage of webservice architecture along with two new major components that is Controller and Agent. The rest of paper is organized as follows , related work on fault tolerance in distributed technologies and Webservices is described in section 2. A basic idea about the technology is presented in section 3. A novel robust and fault tolerance
  • 2. framework using WS-I specification is proposed in section 4, it also includes the description about two major components of proposed framework. Description of proposed framework with step by step presented in section 5. In section 6 conclusion and future work have been described. II. RELATED WORK With the help of related data and information, it is a universally accepted fact that fault tolerance can be achieved with duplicating hardware and software resources [7] [8]. For instance, if services are duplicated and one of them fails, the remaining one can still ensure the desired output. Due to the high demand of web services, issues related to Fault Tolerance are becoming critical. Therefore an appropriate solution is needed to solve such type of issues. In the recent years, several fault handling methodologies and frameworks have been proposed for web services and SOA. But due to a high demand of web services it still requires a more theatrical background and robust and fault tolerant framework to provide high availably and quality of service . Vijay Dialani et al.[9] have proposed a transparent fault tolerance for Web services based architectures which provide capacity for monitoring the individual instance and service host. Chen Liu et al.[10] have proposed a rule-based exception handling approaches in SOA, in which they introduced some basic rules, categories of exception with strategies to handle it. Huang et al.[11] have proposed a runtime monitoring composite web service through stateful aspect extension where they followed stateful aspect extension approach to monitor web service at runtime. Roman Vaculfn et al.[12] have proposed exception handling model with appropriate recovery actions for semantic web services that is purely based on OWL-S. Wen jiajia et al.[13] have proposed a system for handling exception in BPEL process through multi-policy, but proposed work lacks to provide reliability and extensibility of web services. M.G. In [14] author proposed the mechanism of recovering faulty Web apps using Service discovery in which they described the mechanism for handling faults using various kinds of the recovery mechanism. In their work they introduced a self healing architecture for identifying faults and taking appropriate actions based on faults. In [15], service oriented architecture(SOA) based service recovery framework has been proposed. In this work's author introduced an algorithm to select recovery action based on exception servility but it still needs various recovery actions and compatibility based on web service stack of standards such as WSDL, SOAP and UDDI. M. Panahi et al.[16] have defined the a middleware architecture to support an accountable service oriented architecture that consists of three components ASB, AA and Agent. Their work mainly focuses to mange web service faults that to only those are closely related to performance. But in their work various kinds of faults in services are not considered. B. Pernici et al.[17] proposed a work for identifying the fault type and methodologies to fix various types of faults. They categorized the various types of faults based on state such as permanent, transient, which are determined by Bayesian classification. However, their work requires more concrete emphasis to identify the type of fault related to fault duration and criteria to choose methodologies. Tamir et al. [18] proposed a fault tolerance model for webserver which detects server errors and forwards requests to a replica server to avoid service failures. In [19], Author proposed a fault tolerant architecture for grid applications. This model takes a help of a passive replication approach in order to provide fault tolerance. III. TECHNOLOGY BACKGROUND This section contains basic concepts of web services related to the fault handling in distributed technologies and web services. Initially We have applications which are designed to run on one machine. Like when we develop applications in 'C' language , it can execute on a box and can talk to other applications which are running on the same box. But now distributed technology solves the above problems and allows to build applications which is not only executed on a machine and talk to other applications on the same machine, these applications should also be able to access or communicate with other applications which are running on a different machine. The web service is technology that allows us to build an interoperable distributed application component. To build interoperable distributed application component, every software vendor in the industry started defining their own standards. In such case interoperability never will become reality. With a time it was realized by every software vendor that we can't build independently an interoperable distributed application component, unless everyone has agreed upon same standards. There comes a WS-I[6] organization which is formed to build open standards for building interoperable distributed application component. WS-I released BP1.0 specification document, the document contains a set of guidelines or standards that must be used to develop a distributed interoperable component . Web services have three main parts in its architecture [1- 3,5] that is a service provider, service consumer and UDDI registry [21].In order to achieve interoperability in webservices common language for exchange data is used that is XML but having XML is not enough to exchange data between two systems. We need a layer on top of it to classify the information. For instance, We don't send only business data for communication, along with business data, we may need to send transport specific data or some helper data. So using XML we can't differentiate such kind of information. Hence in such cases SOAP is used that can differentiate business data and helper data . In the proposed framework we use a SOAP binding protocol for interaction. This paper focuses in the field of robustness and FT(Fault Tolerance) in webservices. The key idea is to provide a framework that can offer QoS with high availability of web services. Proposed framework uses the existing webservice architecture to ensure compatibility with stack of webservices standards.
  • 3. IV. PROPOSED FRAMEWORK Our framework is based on the existing web service architecture and WS-I organization specification with fault tolerance and robust capability in order to provide QoS and achieve interoperability up to the mark. A. Avizienis et al.[20] given basic concept and dependable and secure computing in which different types of faults are discussed that may arise during service invocation. The proposed framework includes two major components that are responsible to mange replica for diversity actions, fault diagnosis and appropriate recovery actions. In our proposed framework, configuration stage is responsible for publication of web services where the provider of service uses the UDDI registry for registering web service replica. On execution stage, an Agent module is used to create web-service replicas. Once the web - service replica is created Agent performs different test to ensure the service that has to be used. In this stage Controller job is to detect the faults. It also does the verification of the web service during its execution. If a fault gets propagated, the Controller communicates with Agent for selection of appropriate web service that is replicated across various nodes. There are two major modules of proposed framework that is Controller and Agent . 1. Controller: Controller module in framework provides QoS to consumers by taking different kind of necessary actions. It also gives facility to provide for detecting faults and taking actions based upon faults. Figure 1: A Novel Robust & Fault tolerance Framework for Webservices using WS-I* Specification
  • 4. The main responsibility of Controller for fault detection, notification and confinement. It gives notification to the Agent if webservice renders with a fault during the service execution. For example, if webservice gets a fault during its execution, then Controller gives fault notification to the Agent in order to assign appropriate web service replica to continue normal execution from manual breakdown. In the framework Controller gets a request from consumer with an appropriate web service type, including operations and parameters for service execution. Then Controller takes an help of Agent in order to perform mapping of service type in available webservice replica. 2. Agent: The main job of the agent is to manage replicas across the nodes. It creates service replica using UDDI registry [21]. The Agent also checks presence of replica during replica selection stage. V. DESCRIPTION OF PROPOSED FRAMEWORK In the figure 1.0, There can be a separate Controller for each and every web service consumer. In diagram Agent and UDDI can be located remotely at different locations. Here multiple Agent can be deployed in different locations with the Controller. The proposed framework work as presented in following steps:- 1. During execution of application, the consumer sends a request of required type. 2. Consumer request passes to Agent through Controller. 3. Agent searches web-service of required service type from UDDI registry [21]. 4. Now Agent performs an availability test for choosing the appropriate web service replica that has to be used. 5. Next, Agent returns web-service address of the controller. Now the consumer can invoke the needed operation/ method in the selected service through the returned access point. 6. The Controller does the verification of service execution. During execution, if any, faults occur, then controller takes an help of replica and continues the manual execution of the application. 7. Now, when service execution is completed results are reported back to the consumer. Here, the replica creation takes an advantage of UDDI standards [21]. The extended UDDI allows a refined web- service replica selection that is based on the availability, interoperability and response time . VI. CONCLUSION AND FUTURE WORK In this paper a robust and fault tolerant framework for web services using WS-I * specifications is proposed. Our framework uses existing web service architecture, including two more major component Controller and Agent to ensure robustness and fault tolerance in web services. The main benefit of web-service technology is interoperability irrespective of the language and platform through open stack of web service standards such as SOAP, WSDL, XML etc. In future work, the idea is to propose a load balancing algorithm for avoiding from consumer to forward multiple requests to the same web service replica. Further work is also to propose fault tolerance framework for mobile web services in which proposed work may use REST instead of SOAP as smart devices will have very less resources in terms of memory. So by keeping hardware as well as software resources as an issue, the idea is there to use a REST interaction style of web services for mobile devices instead of traditional SOAP. ACKNOWLEDGMENT This work is provisionally supported by KIIT University, Bhubaneswar, India. Our sincere thanks to KIIT University, Bhubaneswar, India for providing resources to complete this work. REFERENCES [1] E. Thomas. SOA Principles of Service Design. Prentice Hall International, 1st edition, 2007. [2] W3C. Simple Object Access Protocol (SOAP) 1.1. http://www.w3.org/TR/soap/. [3] W3C. Web Services Description Language (WSDL). http://www.w3.org/TR/wsdl. [4] OASIS. Universal Description, Discovery and Integration(UDDI). http://uddi.org/. [5] B. Margolis and J. Sharpe. SOA for the Business Developer - Concepts, BPEL, and SCA. MC Press, 1st edition, 2007. [6] WS Interoperability Organization. Web Service Interoperability (WS-I), 2005. [7] B. Kim, “Reliability analysis of real-time controllers with dual-modular temporal redundancy,” Proc. of the Sixth International Conference on Real-Time Computing Systems and Applications (RTCSA)1999, pp.364- 371, 13-15 Dec. 1999. [8] B. Kim, “Reliability analysis of real-time controllers with dual-modular temporal redundancy,” Proceeding of the 6th International Conference on Real-Time Computing Systems and Applications (RTCSA) 1999, pp.364-371, 13-15 Dec. 1999. [9] V. Dialani, S. Miles, L. Moreau, D. D. Roure, and M. Luck. "Transparent fault tolerance for Web services based architectures". In Proceedings of the 8th International Euro-Par Conference on Parallel Processing, pages 889–898. Springer, 2002. [10] Chen Liu, Yanting Xu, D., and Xaiyu, L. “A rule- based exception handlings approach in SOA”. International Conference on Computer Application and System Modeling (2010).
  • 5. [11] Huang T, Wu GQ, Wei J, "Runtime Monitoring Composite Web Services Through Stateful Aspect Extension", Journal Of Computer Science And Technology, vol. 24, pp. 294- 308, 2009. [12] Roman Vaculfn, Kevin Wiesner, Katia Sycara "exception handling and recovery of semantic web services", 4th International Conference on Networking and Services, ICNS 2008, Pages 217-222 ,2008. [13] Wen Jiajia, Chen Junliang, Peng Yong and Xu Meng. "a multi policy exception handling system for bpel process", First Intl. Conf. on Communications and Networking in China, ChinaCom '06, 2007. [14] M.G. Fungini and E. Mussi, “recovery of faulty web applications through service discovery,” Proc. First International Workshop Semantic Matchmaking and Resource Retrieval: Issues and Perspectives (SMR 2006), Sept. 2006, pages. 67-80. [15] C. Jie-Ying, W. Yong-Jun, and X. Yi, “soa-based service recovery framework,” Proceeding the 9th International Conference. Web-Age Information Management (WAIM 2008), July 2008, pages. 629-635. [16] M. Panahi, K.J. Lin, Y. Zhang, S.H. Chang, J. Zhang, and L. Varela, “The LLAMA Middleware Support for Accountable Service-Oriented Architecture,” Proceeding the 6th International Conference Service Oriented Computing (ICSOC 2008), Lecture Notes in Computer Science 5364, Springer- Verlag, Dec. 2008, pp. 180-194. [17] B. Pernici, and A.M. Rosati., “Automatic Learning of Repair Strategies for Web Services,” Proceeding. 5th IEEE European Conf. Web services (ECOWS 2007), IEEE Computer Society, Nov. 2007. Pages. 119-128. [18] N. Aghdaie and Y. Tamir. "Client transparent fault-tolerant web service". In Proceedings of the IEEE International Conference on Performance, Computing, and Communications, pages 209-216. IEEE Computer Society, 2001. [19] D. Zagorodnov, X. Zhang, M. Hiltunen, K. Marzullo and R. D. "Schlichting. Fault-tolerant grid services using primary backup: feasibility and performance". In Proceedding IEEE International Conference. on Cluster Computing, pages 105-114. IEEE-CS, 2004. [20] A. Avizienis, J. C. Laprie, B. Randell, C. Landwehr. Basic concepts and taxonomy of dependable and secure computing. IEEE Transactions on Dependable and Secure Computing, pages 11–33, 2004. [21] UDDI specifications, http://uddi.org/pubs/uddiv3.0.2-20041019.html