Funktionen

Evaluating the Robustness of Data Placement for DRIHM Workflows

Background
The underlying premise of the DRIHM project is that understanding and predicting the environmental and human impact of draughts, floods, landslides and other severe hydro-meteorological events requires a holistic approach. Numerous physical, chemical and biological simulations and datasets describing different aspects of climate change, infrastructure development or land use need to be combined in a way that takes into account interactions between various subsystems. DRIHM provides such an integrated environmental modeling solution, merging collections of several models (in various versions) and numerous data sources (of often unknown provenance) into scientific simulation workflows. While successful simulations deeply depend on combining the right versions of models with the right data in the right environment, the suitability of such simulations for decision making in disaster and emergency management requires in addition a certain degree of robustness in order to guarantee only limited degradation despite fluctuations in the behavior of resources or changes of the environment.

Objectives
In a first step this thesis deals with assessing the robustness of data placements for the DRIHM workflows. Consequently, the objectives of this thesis are fourfold. Firstly, since robust DRIHM workflows would ensure that data requirements are achievable under variable loads, an adequate robustness model needs to be formulated. This thesis will follow the FePIA process and the robustness radius concept. Secondly, given the robustness metric, a set of placement heuristics should be considered for quantifying the robustness of the DRIHM system. Thirdly, the model should be evaluated by experiments in a simulated environment. Finally, the experimental results need to be compared with real observations.

Assigner:
Prof. Dr. D. Kranzlmüller

Requirements:
Basics in Grid Computing, Linux, Python, C++

Duration: according to study guidelines

Number of Students: 1

Supervisors:
Dr. Michael Schiffers, Oettingenstr. 67, Raum E 003 (Erdgeschoß), Tel. 2180-9164