Veriﬁcation On The Web Of Mobile Systems

Gianluigi Ferrari

, Stefania Gnesi

Ugo Montanari

, Roberto Raggi

Gianluca Trentanni

, and Emilio Tuosto

Dipartimento di Informatica, Universit`a di Pisa

ISTI-CNR, Pisa

Abstract. The vast majority of current available veriﬁcation environ-

ments have been built by sticking to traditional architectural style cen-

tralized and without dealing with interoperability and dynamic recon-

ﬁgurability. In this paper we present a veriﬁcation toolkit whose design

and implementation exploit the Web service architectural paradigm.

1 Introduction

The WEB provides uniform mechanisms to handle computing problems which

involve a large number of heterogeneous components that are physically dis-

tributed and (inter)operate autonomously. Recently, several software engineering

technologies have been introduced to support a programming paradigm where

the WEB is exploited as a service distributor.

Our issue in this paper is to demonstrate that the design, development and

maintenance of semantics-based veriﬁcation environments are sempliﬁed by ex-

ploiting WEB services in a modular fashion taking even advantage of the reuse

and integration of “old” modules.

As a preliminary answer to this question a prototype version of a veriﬁca-

tion toolkit has been conceived to support reasoning about the behaviour of

mobile processes speciﬁed in the π-calculus supporting the dynamic integration

of several veriﬁcation techniques.

2 Service Orchestration

The architecture of the toolkit, called Mihda performing minimization of HD-

automata is described in [8], its structure is developed from the co-algebraic

formulation of the partition-reﬁnement minimization algorithm and a web inter-

face is avalaible (http://jordie.di.unipi.it:8080/mihda).

A semantic-based veriﬁcation environment for the π-calculus, called HD Au-

tomata Laboratory (HAL) [1, 2] has been implemented and experimented. HAL

is available at http://fmt.isti.cnr.it:8080/hal.

Work partially supported by FET Projects IST-2001-33100 PROFUNDIS and IST-

2001-32747 AGILE.

Ferrari G., Gnesi S., Montanari U., Raggi R., Trentanni G. and Tuosto E. (2004).

Veriﬁcation On The Web Of Mobile Systems.

In Proceedings of the 2nd International Workshop on Veriﬁcation and Validation of Enterprise Information Systems, pages 72-74

DOI: 10.5220/0002660300720074

 SciTePress

The main issue we have to face consists of making these toolkits accessible

and usable via a WEB interface. This is done into two steps: The ﬁrst step deﬁnes

the WEB orchestration interface which, independently from the implementation

technologies, describes the WEB interaction capabilities; the second step deﬁnes

transforming the program facilities which correspond to publish the orchestration

interface on the WEB.

The main programming construct we exploit to program service orchestration

is XML-RPC to ensures interoperability among components available over the

WEB at the main cost of parsing and serializing XML documents.

In our running example we consider two services, namely the tools HAL and

Mihda. The WEB orchestration interface of Mihda provides three interaction

capabilities: Compile taking a π-calculus agent as input and yielding as output

the corresponding HD-automaton; reduce performing minimization on HD au-

tomata; Tofc2: Transforming the Mihda representation of HD-automata into

the FC2 format used inside HAL. The WEB orchestration interface of HAL pro-

vides three capability: Check performing model checking; unfold generating a

standard automaton out of an HD-automaton; visualize allowing to graphi-

cally operate over HD-automata. The publication on the WEB of the orches-

tration interfaces has been performed by exploiting the facilities of the web

application server Zope providing a comprehensive framework for management

of web contents and mechanisms to ”publish” information on the WEB.

In our experiment, the service orchestration language is python whose ex-

pressiveness gives us the opportunity of programming service orchestration in the

same way traditional programming languages makes use of software libraries. In

particular, services are invoked exactly as “local” libraries and all the issues

related to data marshaling/unmarshalling and remote invocation are managed

by the XML-RPC support. An example of service orchestration is illustrated

below to verify a property of a speciﬁcation, i.e. to test whether a π-calculus

process A is a model for a formula F .

mihda = Server( "http://jordie.di.unipi.it:8080/mihda/hd" )

hal = Server( "http://fmt.isti.cnr.it:8080/hal" )

hd = mihda.compile( A )

reduced_hd = mihda.reduce( hd )

reduced_hd_fc2 = mihda.Tofc2( reduced_hd )

aut = hal.unfold( reduced_hd_fc2 )

check = hal.check( aut, F )

In this few code lines it is clear that the only part of the orchestration that

includes network dependencies is represented by the operation that open the

connections with the HAL and Mihda servers with the service orchestration

program running under WindowsXP, pointing out the interoperability nature of

the toolkit.

However, this network dependency can be removed by introducing a further

module, namely the directory of services together with a simple trader facility

that perform the binding of services that would allow us to avoid specifying the

eﬀective names (and localities) of services into the source code and to dynami-

cally bind services during the execution only on demand, making transparent the

distribution of services and allowing to repicate or to re-allocate a service into a

new locality without requiring any change into service orchestration programs.

3 Lessons Learned

We started our experiment with the goal of understanding whether the WEB

service metaphor could be eﬀectively exploited to develop semantic-based veriﬁ-

cation environments. In this respect, the prototype implementation of a toolkit

supporting veriﬁcation of mobile processes speciﬁed in the π-calculus is a signi-

ﬁcative example.

Our approach adopts a service orchestration model whose main advantage

resides in reducing the impact of network dependencies and of dynamic ad-

dition/removal of WEB services by the well-identiﬁed notions of directory of

services and trader. To the best of our knowledge, this is the ﬁrst veriﬁcation

toolkit that speciﬁcally addresses the problem of exploiting WEB services.

The service orchestration mechanisms presented in this paper, however, have

some disadvantages. In particular, they do not exploit the full expressive power

of SOAP to handle types and signatures. For instance, the so called “version

consistency” problem (namely the client program can work with one version of

the service and not with others) can be solved by types and signatures.

We plan to investigate and experiment the .NET framework to design “next

generation” semantic-based veriﬁcation environments.

References

1. G. Ferrari, G. Ferro, S. Gnesi, U. Montanari, M. Pistore, and G. Ristori. An au-

tomata based veriﬁcation environment for mobile processes. In E. Brinksma, editor,

TACAS’97, LNCS 1217. Springer, Apr. 1997.

2. G. Ferrari, S. Gnesi, U. Montanari, M. Pistore, and G. Ristori. Verifying mobile

processes in the HAL environment. In Proc. 10th International Computer Aided

Veriﬁcation Conference, pages 511–515, 1998.

3. G. Ferrari, U. Montanari, and M. Pistore. Minimizing transition systems for name

passing calculi: A co-algebraic formulation. In M. Nielsen and U. Engberg, editors,

FOSSACS 2002, volume LNCS 2303, pages 129–143. Springer Verlag, 2002.

4. IBM Software Group. Web services conceptual architecture. In IBM White Papers,

2000.

5. U. Montanari and M. Pistore. π-calculus, structured coalgebras and minimal hd-

automata. In Proc. MFCS’2000, volume 1893 of LNCS. Springer, 2000.

6. F. Orava and J. Parrow. An algebraic veriﬁcation of a mobile network. Formal

Aspects of Computing, 4(1):497–543, 1992.

7. M. Pistore. History dependent automata. PhD thesis, Computer Science Depart-

ment, Universit`a di Pisa, 1999.

8. E. Tuosto. Non-Functional Aspects of Wide Area Network Programming. PhD

thesis, Dipartimento di Informatica, Universit`a di Pisa, 2002.