I work within a cooperative project with Fujitsu Technology Solutions.
I'm engaged with the specification of (out-of-band) system management
software of server systems to enhance the quality of the software.
When developing server management software you need to understand the
existing hardware system structures. Especially (out-of-band) system
management software is strongly related to the real hardware system
structures. On lower levels the management software allows hardware
access via hardware-oriented interfaces without having an operating
system installed. On higher levels you find distinct server management
modules that access the lower level components. On each level the
management software can be accessed by so called management interfaces.
You need to understand whether the information and functionality you
want to access is physically wired.
You need to consider the server management software structure as well.
Nowadays, as the server systems become more complex (e.g. a modular
Server System may contains up to several dozens of server nodes) the
system administrator cannot handle the huge number of available modules
manually, so the server management software takes over several
management tasks. E.g. data collection, data consolidation and performs
self-adaptive controlling tasks autonomously. To develop these software
components you need to understand the implicit dependencies between
Based on my industry experience I create a general architecture for
server management software with focus on management functionality to
make implicit structures visible, to give an understandable overview of
the system, and to make development tasks explicit. This knowledge will
be put to a metamodel that extends the Systems Modeling Language
(SysML) e and thus defines an abstract syntax for a specific
architecture description language.
To access management functionality you need to specify the management
interfaces. One uprising standard is the management interfaces based on
the Common Information
Model (CIM), a standard by the
Distributed Management Task Force (DMTF). The specification and
development of these interfaces is a new task for our cooperation
partner and faces some cumbersome, time-consuming tasks. To optimize the
development I create a model-based software development method, which
supports the generation of specification documents, development
artifacts as well as test cases for quality assurance.
Putting all together means to define a software development method for
server management software. This task should be done in a formalized
way. For this I’m using MetaMe, a meta-method for the definition of
software development methods. One major flaw is the non-existing
consideration of situational factors. Situational factors describe the
situational context of the software development method, thus influences
and constraints the method faces during execution. My idea is to capture
all situational factors in advance and set up appropriate method
requirements, which need to be considered during method development.
This approach should help to define a method that fits to the identified
situation. I evaluate the extended MetaMe approach on my industrial
- Zeitschriftenbeiträge (2)
Masud Fazal-Baqaie, Baris Güldali, Markus Luckey, Stefan Sauer,
Michael Spijkerman: Maßgeschneidert und
werkzeugunterstützt Entwickeln angepasster Requirements
Engineering-Methoden. In Dr. Ernst Sikora
(eds.): OBJEKTspektrum (Online Themenspecials), no. RE/2013, pp.
1-5. SIGS DATACOM (2013)
Michael Spijkerman, Tobias Eckardt: Modellbasiertes
Testen auf Basis des fundamentalen Testprozesses. In
Softwaretechnik-Trends, vol. 29, no. 4. GI
- Rezensierte Workshopbeiträge (1)
Michael Spijkerman: Ein pragmatischer Ansatz zur
Entwicklung situationsgerechter Entwicklungsmethoden. In
M. Kuhrmann, D. M. Fernández, O. Linsen, A. Knapp (eds.):
Software Engineering 2013 - Workshopband (inkl.
Doktorandensymposium) SE 2013 (to appear). GI, LNI, vol. 215,
pp. 425-434 (2013)
- Technische Berichte (1)
Matthias Gehrke, Baris Güldali, Jan Meyer, Yavuz Sancar, Michael
Spijkerman: “Metriken” – Kennzahlen zur
Verbesserung der Softwarequalität, Teile 1 und 2. s-lab
report, no. 3. University of Paderborn, s-lab (to be