Abstract-Band - Fakultät für Informatik, TU Wien - Technische ...

search history we propose a strategy that directs the search to move quickly
through regions, in which better trees are unlikely found. Furthermore, we
present a vectorization technique to speed up computations in the tree
evaluation function. Our results showed that IQ-Tree runs two to four times
faster than IQPNNI, while keeping the accuracy of the reconstructed
phylogenies.
Arbeitsbereich Distributed Systems
Gabriel Kittel
Survey and Taxonomy of Autonomic Large-scale Computing
Studium: Masterstudium Software Engineering & Internet Computing
Betreuer: Univ.Prof. Dr. Schahram Dustdar
In the area of distributed systems, several approaches have emerged with the
objective to deliver computing power as a public utility like water, gas,
electricity and telephony. Cloud computing is the latest of those approaches
where virtual computing infrastructure, software development platforms and
applications are provisioned on demand over the Internet. Autonomic
computing is a computing paradigm that promises to deliver systems
adapting themselves to environmental changes by employing selfmanagement
mechanisms guided by policies as an eort to address the
management complexity that arises from the dynamics of resource availability
and system load in large-scale computing systems. The Foundations of Selfgoverning
ICT Infrastructures project (FoSII) at TU Vienna intends to enhance
self-management support in existing service-oriented architectures. A survey
of existing projects that introduce autonomic computing to large-scale
computing systems like grids and clouds, and a taxonomy that provides
classication criteria for that research eld would allow to assess the current
state of research by suggesting criteria to help identify application areas,
subproblems and approaches within that eld. However, such a survey and
taxonomy have not yet been proposed to this day. The goal of this thesis is to
systematically investigate the state of art of self-management by providing a
taxonomy of autonomic large-scale computing. It presents a survey of
projects and theoretical work in that eld and proposes a taxonomy that
classies autonomic large-scale computing. Survey and taxonomy allow to
assess the current state of research in autonomic large-scale distributed
systems, thus supporting further advancements in the eld of autonomic largescale
distributed computing.
64