=Paper=
{{Paper
|id=Vol-1838/paper-08
|storemode=property
|title=Towards Applying the Normalized Systems Theory to IT Infrastructure Systems
|pdfUrl=https://ceur-ws.org/Vol-1838/paper-08.pdf
|volume=Vol-1838
|authors=Geert Haerens
}}
==Towards Applying the Normalized Systems Theory to IT Infrastructure Systems==
https://ceur-ws.org/Vol-1838/paper-08.pdf
Towards Applying the Normalized Systems
Theory to IT Infrastructure Systems
Geert Haerens
Antwerp University, Antwerp, Belgium, geert.haerens@student.uantwerpen.be
Engie, www.engie.com, geert.haerens@engie.com
Abstract. The agile enterprise requires evolvability at all layers - busi-
ness, application and infrastructure. IT infrastructure systems are the
foundations of IT systems in general. Their evolution has a profound
impact on applications and business capabilities. Normalized Systems
Theory (NS) provides a method to evaluate the evolvability of modular
systems. As IT infrastructure systems can be represented by a modular
structure, NS can be used to study their evolvability. An artefact is be-
ing proposed, made up of a 4-step method and summary table, to study
the evolvability of a modular system representing an IT infrastructure
system by means of NS. Although the artefact has been successfully ap-
plied to some IT infrastructure systems and as such demonstrating NS
can be applied to IT infrastructure systems, the 4-step method requires
refinement and a more rigorous translation of the NS theorems into IT
infrastructure equivalents. Further research on the subject is being pro-
posed.
Key words: IT infrastructure systems, Normalized Systems, modular-
ity, evolvability
1 Introduction
The agile and morphogenic enterprise [1] requires the capability to cope with
constant change. Business organization, processes, applications and infrastruc-
ture need to cope with the required agility.
No application will run without IT infrastructure. IT infrastructure components
such as CPU, memory, operating system, are the bare necessities to run any
application. Changes and evolutions in the IT infrastructure can lead to serious
Combinatorial Effects - hidden coupling or dependencies in a system which in-
crease with the size of the system inducing a ripple effect throughout the whole
IT landscape and breaking other IT infrastructure components, applications and
even business processes. Architecting and constructing IT infrastructure systems
which are resilient to those changes and have proven evolvability, are as impor-
tant as the creation of applications and business processes which have proven
1
Copyright by the paper’s authors. Copying permitted for private and academic
purposes. In: Aveiro et al. (Eds.): Proceedings of the EEWC Forum 2017, Antwerp,
Belgium, 09-May-2017 to 11-May-2017, published at http://ceur-ws.org
2 Geert Haerens
evolvability. They are an integral part of the agile and morphogenic enterprise.
Normalized Systems (NS) ( see [2] [3] [4] [5]) provides theorems to evaluate the
evolvability of modular structures. Using design science, a 4-step method has
been worked out to apply the NS theorems to a modular representation of an IT
infrastructure system. The artefact has been applied to several IT infrastructure
systems, demonstrating the feasibility to apply NS on IT infrastructure systems.
2 Proposal
Design science identifies a method, a step approach to address a problem, as a
valid artefact to apply to a problem (see [6], [7],[8]). Applying the NS theorems
to an IT Infrastructure system has been translated into a method to check the
compliance of a relevant modular representation of an IT infrastructure system
with the 4 NS theorems, being:
1. SoC : Separation of Concerns
2. SoS : Separation of State
3. VT : Version Transparency
4. IT : Instance Traceability
The proposed artefact contains the following steps:
– Step 1: Create a relevant modular representation of the IT Infrastructure
system.
– Step 2: For each module of the modular representation look for manifestations
of Concern, State, Version and Instance.
– Step 3: Check if the manifestations found in Step 2 are compliant with the 4
NS principles.
– Step 4: If there is non-compliance with one or more of the 4 NS principles,
describe related Combinatorial Effects (CE).
The results of the 4 steps can be represented in the following summary table:
Fig. 1. Artefact summary table
NS and IT infrastructure 3
The artefact can also be used in the opposite direction, meaning that based
on observed Combinatorial Effects (CE), the violation of one or more of the NS
principles can be identified and this violation can be associated to a manifesta-
tion of Concern, State, Version and Instance in a module representing a function
and/or construction component of an IT infrastructure system.
3 Evaluation
In [9] the artefact has been applied on 3 IT infrastructure systems, being Hous-
ing (data center setup), Hosting (server hardware and an Operating System)
and Proxy (network proxy for N to 1 outbound network traffic). For Housing,
Hosting and Proxy, respectively 3 (Housing 1.0, 2.0 and 3.0), 2 (Hosting 1.0
and 2.0) and 1 implementation patterns have been investigated, resulting in 6
effective use cases on which the artefact has been applied.
These 6 use cases have been evaluated by an expert team (13 members) which
represent the Knowledge Base in the Design Science framework of Paul Johan-
nesson and Erik Perjons [6]. The expert team has been requested to evaluate
the correctness of the used IT infrastructure modular structure (step 1 for each
of the 6 uses cases), the correctness of the analyses (steps 2 to 4 for each of the
6 uses cases), and score the relevance of the artefact (did it provide additional
insight for each of the 6 uses cases). The results can be found in Fig 2.
Fig. 2. Artefact evaluation results
4 Geert Haerens
4 Conclusion and further research
The previous section shows that the proposed artefact can be used to apply NS
on IT infrastructure system, and can thus be used to test evolvability of the IT
infrastructure system. However, the different steps of the artefact method do
require extensive knowledge of the IT infrastructure system under investigation
and a good understanding of what the manifestations of Concern, State, Version
and Instance may look like in the IT infrastructure system.
Further research is required to create a standardized meta model which can be
used to make a modular representation of an IT infrastructure system. A deeper
understanding of the meaning of Concern, State, Version and Instance in an
IT infrastructure system needs to be further investigated. Based on this deeper
understanding, the transformation of the 4 NS theorems into applicable Nor-
malized Infrastructure Systems Theorems (NIST) can be created. The current
artefact summary table must be improved by having a standardized way to de-
scribe manifestations of Concern, State, Version, Instance and the description
of the CE. The proposed artefact improvements will lead to a more systematic
approach in applying the artefact.
Once a transformation of the 4 NT theorems into applicable Normalized In-
frastructure Systems Theorems (NIST) is available, the research can shift to-
wards using those as input for an expander which will, based on a standardized
functional description of an IT infrastructure system, expand code which can
be deployed on a Software Defined Infrastructure Platform (like AWS, Azure,
Google), resulting in deployable and evolvable IT infrastructure systems.
References
1. Jan A.P. Hoogervorst - Enterprise Governance and Enterprise Engineering - 2009
2. Herwing Mannaert, Jan Verelst - Normalized Systems - 2009
3. Herwig Mannaert, Jan Verelst, Peter De Bruyn - Normalized Systems Theory: From
Foundations for Evolvable Software Toward a General Theory for Evolvable Design
- 2016
4. Herwig Mannaert, Jan Verelst, Kris Ven - The transformation of requirements into
software primitives: Studying evolvability based on systems theoretic stability -
Science of Computer Programming: Volume 76, Issue 12 pp. 1210...1222 - 2011
5. Herwig Mannaert, Jan Verelst, Kris Ven - Towards evolvable software architectures
based on systems theoretic stability - Software Practice and Experience: Volume 42,
Issue 1, pp. 89...16 - 2012
6. Paul Johannesson, Erik Perjons - An Introduction to Design Science - 2014
7. Hevner, A. R. and March, S. T. and Park, J. and Ram, S. - Design Science in
Information Systems Research - MIS Quarterly: Volume 38, Issue 1 pp. 75..105 -
2004
8. Peffers, K. and Tuunanen, T. and Rothenberger, M.A. and Chatterjee, S. - A De-
sign Science Research Methodology for Information Systems Research - Journal of
Management Information Systems: Volume 24, Issue 3, pp. 45..77 -2007
9. Geert Haerens - Master Thesis: Applying the Generalized Normalized Systems The-
ory to the Engie IT Reference Architecture Library - 2016