ROA MODULAR LDAP-BASED APPROACH TO INDUSTRIAL
SOFTWARE REVISION CONTROL
Cristina De Castro
IEIIT-CNR, Italian National Research Council
V.le Risorgimento 2, Bologna 40136, Italy
Paolo Toppan
CNIT, Italian Inter-University Consortium for telecommunications,
V.le Risorgimento 2, Bologna 40136, Italy
Keywords: Software revision control, industrial context, plant installations, LDAP schemes for revision
control.
Abstract: A software revision control sy
stem stores and manages successive, revised versions of applications, so that
every design stage can be easily backtracked. In an industrial context, revision control concerns the
evolution of software installed on complex systems and plants, where the need for revision is likely to arise
from many different and correlated factors. In this paper, starting from the wide bibliography available on
the subject, some typical schemes are discussed for representing such factors. An LDAP-based architecture
is addressed for modelling and storing their evolution.
1 INTRODUCTION
The efficient management of applications’ lifecycle
is becoming more and more important and complex
in industrial environments, where software is
installed in large and often distributed plants. The
overall process results from several different factors:
for instance, a malfunctioning can arise from the
incorrect interaction between an old software
version running on an updated or modified machine;
software results generally from the synergy of many
different people, approaches and requirements, etc.
The problem of revising software in industrial plants
is thus more and more challenging and the critical
issue of software documentation is bound to become
more and more essential in the backtracking process.
Versioning is widely used for the management of
software
development and revision (Westfechtel et
al., 2001, Fischer et al, 2003), even if not always
with a precise temporal semantics. This process,
named revision control, allows to backtrack and
compare every design stage. Far from expecting to
solve this very difficult task, this paper discusses
some simple, typical schemes for representing a
software revision process and proposes an LDAP-
based temporal architecture for their representation.
2 COMPONENTS OF AN
INSTALLATION, EVOLUTION
AND THE LDAP SCHEME
In an industrial context, software can not be
considered an isolated issue. As a matter of fact,
software is installed on a set of machines and has
related configuration specifications. An installation
instance (Fig. 1) can thus be defined as the set of
these three groups of components and their
interaction. In this scheme, the plant requirements
and the documentation of software, hardware and
installation features, detailing all the design phases
and the interactions, are considered part of the
installation instance. As a matter of fact, they play a
fundamental role in the process of software
development and revision. An installation instance
represents the snapshot of the state of a plant after a
generic installation or update, thus it represents a
version in the process of industrial software revision
control. Let us now consider the generic structure of
an installation instance, one of its versions is the
collection of the following component versions: (1)
requirements version; (2) software version; (3)
hardware version; (4) configuration files version; (5)
documentation version.
261
De Castro C. and Toppan P. (2006).
ROA MODULAR LDAP-BASED APPROACH TO INDUSTRIAL SOFTWARE REVISION CONTROL.
In Proceedings of the First International Conference on Software and Data Technologies, pages 261-265
Copyright
c
SciTePress
Figure 1: Installation instance.
In order to allow a component version to be shared
and referred by many installation instances, each of
the components can be labelled independently. As
far as the type of temporal labelling is concerned, it
is quite common to label the versions by means of
progressive numbers, but the necessity can arise of
referring the system to a precise temporal phase of
its lifecycle. For instance, if a malfunctioning causes
trouble to a client, it can be of great importance,
even from a juridical point of view (
Grandi et al,
2003, Harris et al, n.d.), to know precisely how the
installation was when the problem occurred. The
support of at least two kinds of time dimensions is
widely emphasized in literature (
Böhlen et al, 2006,
Elmasri et al, 1990, Tansel et al, 1993): transaction
time, which tells when an event is recorded or
updated in a knowledge base, and valid time, which
represents when an event occurs, occurred or is
expected to occur in the real world. Transaction-time
is represented by adding the endpoints IN, OUT and
valid-time is represented by the endpoints FROM,
TO. In the considered environment, transaction time
tells when the installation version description
discussed above was recorded in the revision control
knowledge base and valid time refers to when the
installation is operative in the plant. For instance, the
installation version h in Fig. 2 was recorded in the
knowledge base on May 3rd 2006 and it has not
been updated yet, so IN = May 3rd 2006 and OUT =
∞. It has been fully operative on the plant FROM
May 15th 2006 and is expected to be used until (TO)
December 23rd 2007. Note that there is no need a
priori that they should be labelled with the same
values of transaction and valid-time. For instance,
the software version j can have been recorded before
May 3rd 2006 but used within the installation
version h. In this way, components can be shared
and duplication is partly avoided. For instance, the
installation version g in the background in Fig. 2
shares software version j.
A further advantage of using this kind of
timestamping is that, if a bug is detected within an
installation version, all the other versions that may
be touched by the same bug can be found in a very
simple way by means of temporal queries.
This structure can be easily developed by means of
the Lightweight Directory Access Protocol (LDAP).
LDAP (
Howes et al, 2003, Koutsonikola et al, 2004) can
be particularly suitable for representing installation
versions in wide enterprises: as a matter of fact,
LDAP is widely used in enterprise databases and it
is optimised for reading operations, being thus
suitable for storing and managing temporal data that
must be backtracked. Furthermore, LDAP schemes
can be very easily changed and extended in order to
record new attributes and new classes. This feature
can be very useful when designing a revision
control. As a matter of fact, new objects and new
attributes are very likely to be modified or added due
to new needs or improvements made by the people
who develop it. LDAP was also built for distributed
environments, so it suits the distributed location of
industrial applications very well.
Referring to the schemes discussed above, the
LDAP object-classes tree can be defined as follows
(see Fig. 3, Tab. 1): the 0-level class describes the
installation in general; the 1-level classes are
respectively plant requirements, software, hardware,
configuration files and documentation. It must be
observed that the timespan of the different objects
can have different meanings. For instance, the
temporal attributes of class “software” have the
following meaning: 1) IN, OUT: when the software
version was recorded/updated in the revision
information system; 2) FROM, TO: the period in
which the software version is/was/will be operative.
ICSOFT 2006 - INTERNATIONAL CONFERENCE ON SOFTWARE AND DATA TECHNOLOGIES
262
Figure 2: Timestamps of an Installation Version and its components.
Figure 3: An LDAP basic structure for the representation of installation versions.
3 CONCLUSIONS
In this paper, a basic architecture was discussed for
representing the temporal evolution of a software
installation on an industrial plant. This model
divides an installation in many interacting
components and, making use of the LDAP model,
represents them in a hierarchy. The evolution of the
system is represented by means of
both transaction-time and valid-time. Future work
will be devoted to deepen the semantics and the
interaction mechanisms of the components of this
model, in order to merge it in a revision control
system.
ACKNOWLEDGEMENTS
The authors want to thank the InSeBaLa project of
Regione Emilia Romagna (Italy) for supporting this
work.
ROA MODULAR LDAP-BASED APPROACH TO INDUSTRIAL SOFTWARE REVISION CONTROL
263
Table 1: Semantics of classes and attributes.
Class Attributes Description
installation oid, name, description
plant
IN, OUT
FROM, TO
the specific industrial plant where the installation
was made
when the installation version was
recorded/updated in the revision information
system
the period in which the installation version
is/was/will be fully operative on the plant
plant
requirements
oid , plant, goals (multivalued), description,
technical requirements (multivalued)
….
IN, OUT
FROM, TO
when the requirements version were
recorded/updated in the revision information
system
the period in which the requirements
are/were/will be the actual reference of the plant
software oid, name, description, plant , goals (multivalued)
hardware requirements (multivalued)
….
IN, OUT
FROM, TO
it can refer to the hardware version/s of
another/others installation version/s
when the software version was recorded/updated
in the revision information system
the period in which the software version
is/was/will be operative on the plant
hardware oid, name, description, plant, goals (multivalued),
software requirements (multivalued)
….
IN, OUT
FROM, TO
it can refer to the software version/s of
another/other installation version/s
when the hardware version was recorded/updated
in the revision information system
the period in which the hardware version
is/was/will be operative on the plant
configuration
files
oid
name, description, plant, goals (multivalued)
IN, OUT, FROM, TO
similar semantics
documentation oid , plant, description, goals, …
software description, hardware description
IN, OUT, FROM, TO
it can refer to the software version/s of
another/other installation version/s
….
ICSOFT 2006 - INTERNATIONAL CONFERENCE ON SOFTWARE AND DATA TECHNOLOGIES
264
REFERENCES
Böhlen, M., Gamper, J., Jensen, C.S. (2006). “Multi-
dimensional aggregation for temporal data”. Proc. of
EDBT-2006, pp. 257-275.
Elmasri, R., El-Assal, I., Kouramajian, V. (1990).
“Semantics of Temporal Data in an Extended ER
Model”. Proc. Intl. Conf. on E-R Approach, pp. 52-61
Fischer, M., Pinzger, M., Gall, H. (2003). “Populating a
Release History Database from Version Control and
Bug Tracking Systems. Proc. of 19th IEEE
International Conference on Software Maintenance
(ICSM'03), pp. 23-29.
Grandi, F., Mandreoli, F., Tiberio, P., Bergonzini, M.
(2003). “A Temporal Data Model and Management
System for Normative Texts in XML Format”. Proc.
of WIDM’03.
Harris, C., Allen, R. B., Plaisant, C., Shneiderman, B.
(n.d.). “Temporal Visualization for Legal Case
Histories”. http://hcil.cs.umd.edu/trs/99-18/99-18.html
Howes, T. A., Smith, M. C., Good, G. S. (2003).
Understanding and Deploying LDAP Directory
Services, Addison Wesley 2003, 2^ ed.
Koutsonikola, V., Vakali, A. (2004). “LDAP: framework,
practices, and trends”, IEEE Internet Computing, Vol.
8, Issue 5, Pages: 66 - 72
Tansel, A., Snodgrass, R.T., Clifford, J., Gadia, S.K.,
Segev, A. (eds.) (1993). Temporal Databases, Theory,
Design and Implementation. Benjamin-Cummings.
Westfechtel, B., Munch, B. P., Conradi, R. (2001). “A
Layered Architecture for Uniform Version
Management”, IEEE Transactions on Software
Engineering, Vol. 27, N. 12
ROA MODULAR LDAP-BASED APPROACH TO INDUSTRIAL SOFTWARE REVISION CONTROL
265
