Author: Marcus Hilbrich

We are delighted that Humboldt-Universität decided to fund a new 
Interdisciplinary Center on Greencompute – saving 
energy in large-scale data analysis, led by Ulf Leser (Computer 
Science) and Claudia Draxl (Physics).

Our next Retreat is on the 9th and 10th of November at Seepark Hotel am Wandlitzsee.

See you there.

FONDA invites you to the presentation Dynamic Distributed Physical Design for Database Systems by Khuzaima Daudjee (University of Waterloo) on 27th June 2023, 15:00. The presentation is at Großer Seminarraum (Room 219/220), Albert-Einsteinstrasse 9, Berlin (Adlershof).

Abstract:

Database workloads can be varying, placing changing demands on a system. To meet this challenge, a distributed database system can adjust its physical design, i.e., its data mastering, replication and partitioning schemes to deliver good performance.  I will present MorphoSys, a distributed database system we have built that caters to workload shifts and changes through autonomous metamorphosis of its distributed physical design.

Bio:

Khuzaima Daudjee is a professor at the University of Waterloo.  His research interests are in designing and building large-scale data systems and on providing systems-level support for applications such as streaming, graph processing and machine learning.  His work has led to several awards including Best Paper Award at ACM Symposium on Cloud Computing and ACM SIGMOD Best Demonstration Awards.  He is an ACM Distinguished Scientist.

Special thanks to Prof. Dr. Ina Schaefer (Karlsruher Institut für Technologie) for the presentation on “Correctness-by-Construction – How Can We Build Better Software?” and to Prof. Dr. Thomas Kosch (HU-Berlin) on “ChatGPT und Worfklows”.

The Collaborative Research Center 1404 FONDA (Foundations of Workflows for Large-Scale Scientific Data Analysis) welcomes the first two guests to our female internship program.

Annaëlle Maire is a master’s student at the Télécom Physique Strasbourg
engineering school, studying computer science. As part of her end-of-study internship, she joined FONDA to work on current workflow management systems. More specifically, she is working on FONDA’s B5 sub-project and is participating in the development of a common API between a workflow management system and a resource manager in order to optimize the scheduling of workflows.

Pauline Karega is a master’s student from Nairobi, Kenya. She is
interested in genomic data science and is studying genomic workflows to
help in subproject A2 in FONDA.

The next FONDA Retreat will take place from June 5 to June 6.
We will go to Tagungshotel Sommerfeld in Kremmen this time (link to Openstreetmap).

For poster templates and additional information, visit our internal wiki.

For any further questions, contact our Scientific Coordinator.

Keynote

Prof. Dr.-Ing. Ina Schaefer (Karlsruher Institut für Technologie)

Correctness-by-Construction – How Can We Build Better Software?

Correctness-by-Construction (CbC) is an incremental software development technique to create functionally correct programs guided by a specification. In contrast to post-hoc verification, where the specification and verification mainly take part after implementing a program, with CbC the specification is defined first, and then the program is successively created using a small set of refinement rules. This specification-first approach has the advantage that errors are likely to be detected earlier in the design process and can be tracked more easily. Even though the idea of CbC emerged many years ago, CbC has not yet achieved its full potential in industrial practice. However, we believe that a scaled CbC approach contributes to solving problems of modern software development. In this tech talk, I will give an overview of our work on CbC in four different lines of research, including developing configurable as well as safety- and security-critical software systems. For all of these, we provide tool support building the CorC ecosystem that facilitates CbC-based development in different fields of application and for differing sizes of software systems.

Time Plan

Welcome to our IRTG on June 13th!

The IRTG takes place on Tuesday the 13th of June 2023 in  Erwin
Schrödinger-Zentrum, Rudower Chaussee 26, 12489 Berlin (Adlershof),
Vortragsraum 0’101.

Afterward, we will go over to the MOPS (link to the location) again to have a BBQ together.

The workshop takes place this Thursday, and Friday, 16th/17th March 2023. The location is Humboldt-Kabinett at Institute for Computer Science in Adlershof, Rudower Chaussee 25, 12489 Berlin.

On Thursday, we start with a Tutorial on DSL by Markus Völter. From 15:00 until 18:00.

On Friday, we continue with the workshop at 9:00 and end at about 16:00. The lunch is self-paid.

The time plan for Friday is the following:

09:00 - 09:30 	Ulf Leser
	Introduction, Welcome, Get together

09:30 - 10:30 	Bernhard Rumpe (Aachen)
	Keynote

10:30 - 10:45 	Break

10:45 - 11:05 	Reiner Jung (Kiel University)
	CP-DSL: Configure and Setup Scientific Models

11:05 - 11:25 	Somayeh Mohammadi (FU Berlin FONDA A2)
	Using DSL to rewrite bioinformatic DAWs

11:25 - 11:40 	Break

11:40 - 12:00 	Evgeny Kusmenko (Infineon)
	MontiAnna: A Model-Driven Approach to Machine Learning

12:00 - 12:20 	Jan Philipp Albrecht (HU Berlin, FONDA A5)
	A Framework for Systematic Benchmarking of Transfer Learning Approaches for Biomedical Imaging

12:20 - 13:30 	Lunch

13:30 - 13:50 	Markus Scheidgen (HU Berlin / FAIRmat)
	Enabling Material Scientists to Define, Provide, and Use Research Data

13:50 - 14:10 	Sebastian Müller / Martin Kuban (HU Berlin, FONDA A3)
	Fuzzing Computational Materials Science Code Parsers

14:10 - 14:25 	Break

14:25 - 14:45 	Sven Gundlach (Kiel University)
	Test-Driven Development DSL for Ocean Modeling

14:45 - 15:05 	Nico Jansen (RWTH Aachen)
	From Data Models to Information Systems

15:05 - 16:00 	Discussion

16:00 	Farewell 

We have one more Lecture Series this semester:

“Software Engineering for Science and Science for Software Engineering”

We are happy to have Andreas Zeller (CISPA Helmholtz Center for Information Security) and Norman Peitek (Saarland University) as speakers.

The Lecture Series is on 15th February (15:00 – 18:00) at Berlin Adlershof, Rudower Chaussee 26, Room 1’305. (Not our usual room!)

The first presentation is

“Effective Notebooks: Making Notebooks Reusable, Extensible, and Well-Tested” (Andreas Zeller)

Abstract:

Notebooks combine code, documentation, tests, visualizations, and interactive tutorials in an elegant and reusable fashion, and have quickly become the tool of choice for any researcher exploring, evaluating, and managing large amounts of data. Yet, notebooks are not as reusable and replicable as they could be – and as they should be. In this talk, I highlight some of the problems researchers face when they want to reuse, extend, or replicate notebook code and results, and I introduce several best practices from Software Engineering that improve modularity, quality assurance, documentation, and configuration management – and thus vastly facilitate reuse and extensibility of notebooks.

The second presentation is

“Neuroscience & Software Engineering: A New Wave of Research on Programmers” (Norman Peitek)

Abstract:

Software engineering has traditionally focused on technical aspects such as paradigms, processes, and practices. But, a major component to success—programmers and their minds—has eluded our comprehensive understanding despite decades of research. In the last few years, neuroscientific methods have opened a new perspective for software engineering. In this talk, we provide an overview of the insights and challenges of measuring programmers’ brains.

The Lecture is hosted by Lars Grunske. In case of “technical” questions contact Sebastian Müller and Jan Arne Sparka.

As always: You are all cordially invited!

On 26th January we meet at Albert-Einsteinstrasse 9, Berlin (Adlershof), Großer Seminarraum (Room 219/220).

We are happy to have Vojtech Cima (IT4Innovations – National Supercomputing Center) and Carl Witt (KNIME AG) as speakers.

The time plan is the following:

• 10:00 Introduction to FONDA (Ulf Leser)
• 10:10 Workflow execution at large-scale distributed environments (Vojtech Cima)
• 11:05 Break
• 11:25 Predictive Resource Management for Scientific Workflows (Carl Witt)
• 12:20 Scheduling in Fonda (Fabian Lehmann, Jonathan Bader, Svetlana Kulagina)
• 13:00 Lunch together

You are all cordially invited!

(The next Lecture Series is on 15th February on “Software Engineering for Science and Science for Software Engineering”)