ZIH-Info No. 181, July/August 2024

Extended functions in Datashare

In addition to direct releases to individual logins, self-created groups can now also be configured as targets in the ZIH Datashare service. The prerequisite for this is that the groups of people intended for use have previously been linked to the Datashare service in the group administration of the self-service portal. Another new feature is the option to integrate external storage into the own cloud. The S3 object storage and group drives operated by ZIH are currently supported. Further information: https://tu-dresden.de/zih/dienste/service-katalog/zusammenarbeiten-und-forschen/datenaustausch/cloudstore (Contact: , HA: 40000)

Firewall

TU Dresden's perimeter firewall (previously known as the central firewall / ZFW) was replaced on July 2, 2024 with new devices from the current firewall framework agreement (see ZIH-Info No. 173 and 175). The new firewall system consists of four Palo Alto Networks (PA-5420), which form a high-availability cluster distributed across the two data center locations of ZIH. Each device offers a throughput of around 70 Gbit per second. The new firewalls have integrated application detection regardless of the port/transport protocol used. By using regularly updated virus and malware data, malicious data packets can be detected. The use of a routed connection of the firewalls via the BGP routing protocol enables a much more flexible and reliable integration into the TU Dresden data network (Contact: , HA: 40000).

Further development of the self-service portal

The range of functions of the self-service portal has been further expanded: As a new module, applications for activations in the central firewall have been added, replacing the previous PDF form. In addition, the creation of a group of identically configured mailboxes for several rooms or resources in the Exchange has been simplified by allowing existing mailboxes to be cloned. When applying for email certificates, users can now create a certificate for the functional accounts they manage. There are also new features in the Enterprise Cloud: for example, users are now informed about the upcoming expiry of virtual machines. In addition, the respective head of the organization is automatically pre-filled for the various applications (Enterprise Cloud VMs, Sharepoint sites, firewall shares) as a convenience function. When processing requests, approvers can also see directly who has processed the previous steps and in which role (e.g. as department head or IT Advisor). This function is to be expanded so that changes in the processing process will also be visible in the future. To ensure the sustainability of developments in the self-service portal, we have updated several central libraries to the latest versions in the background. Link to the portal: https: //selfservice.tu-dresden.de (Contact: , HA: 40000)

IT infrastructures for Collaborative Research Centers

On 29 May 2024, the DFG approved the funding of three Collaborative Research Centers (CRC) / Transregio (TRR) for the TUD: the new CRC / TRR 393 "Trajectories of Affective Disorders" as well as the second funding phase of the CRC / TRR 280 "Design Strategies for Material-Minimised Carbon Reinforced Concrete Structures" and the CRC 1415 "Chemistry of Synthetic 2D Materials". The ZIH is involved in all three projects with so-called information infrastructure projects (INF). In these projects, IT infrastructures are being set up for the researchers, here from Medicine, Civil Engineering and Chemistry, in order to ensure FAIR data management, easier data collection and sharing and long-term archiving of research data (Contact: Dr. Ralph Müller-Pfefferkorn).

Expansion of the data trust office for Saxony

In the "DDtrust" project, TU Dresden and other partners have been developing and testing models of a data trust office for the exchange and provision of data for and from research since 2022 - with stakeholders from research and industry; regardless of discipline or data type. The main objectives of the now approved, BMBF-funded follow-up project "DDtrust-scale" are the scaling - i.e. the expansion and establishment - of the data trust office, including the legally compliant design of the processes and the business model as well as the transfer to operations. Through the analysis and practical implementation of further use cases from various disciplines, the processes and basic rules are to be further developed, legal templates created and technical support expanded. Based on the preliminary considerations, the business model is to be worked out in concrete terms, designed in a legally compliant manner and implemented. Under the leadership of TUD's CDIO, the ZIH is working together with partners including TUD|excite, IRGET and UTM. (Contact: Dr. Ralph Müller-Pfefferkorn).

Competence network data trust models

Scientists at TU Dresden are working with partners from Hesse, Saxony and North Rhine-Westphalia to establish a "Competence network data trust models". Under the leadership of TU Darmstadt, the network submitted an implementation concept for the BMBF call for proposals "Scaling and increasing the acceptance of intersectoral data trust models in practice", which was approved for funding. After TU Dresden has already successfully cooperated in the joint projects "DDtrust" for the Saxon scientific area (2022 to 2024) and "DDtrust-scale" (2024 to 2025, see previous article) on the conception, development and scaling of such models, an umbrella structure is now to be created through the "Competence network data trust models". The aim of this project is not only to build up expertise as well as pool skills and networking, but also to pursue the practical and strategic further development of the data trust concept. The project will run for 25 months and is funded by the BMBF with around EUR 2 mill. from European Union - NextGenerationEU funds. The TUD is involved under the leadership of the CDIO, CIDS, TUD|excite and IRGET. (Contact: Prof. Dr. Lars Bernard; CDIO)

DaFne platform: data fusion generator

The BMBF-funded project for the development of the DaFne platform was successfully completed in February. The open source platform is used to generate and evaluate synthetic data to support data-driven decisions in research and industry. It addresses the lack of available data, especially in areas with sensitive information, by generating synthetic, high-quality data. The DaFne platform offers functions such as rule-based data creation, data fusion and reproduction and is geared towards different data types such as numbers, categories, timestamps and geodata. With its focus on smart cities, DaFne provides useful data for applications in the fields of urban planning, finance, medicine, etc. Further information: https://scads.ai/transfer-2/projects/dafne-project/ (Contact: Dr. Iryna Okhrin)

Two ZIH colloquia in August

On Thursday, August 22, at 15:00, a ZIH colloquium with Johannes Langguth from the Simula Research Laboratory, Norway, will take place in Willers-Bau A 317 on the topic "Graph Algorithms on Emerging Tile-Centric Accelerators". Mr. Langguth will present current hardware aspects and programming methods for hardware accelerators specifically designed for machine learning applications.
A special colloquium with Prof. Dr. Stefan Schuster and Shalu Dwivedi from the Friedrich Schiller University Jena entitled "Using game theory to elucidate microbial and cancer pathogenesis" will take place on Thursday, August 29, at 15:00 in the Andreas-Pfitzmann-Bau APB 1096. Further Informationen: https://tu-dresden.de/zih/kolloquium. (Contact: Dr. Hartmut Mix)

Further events

04.07.2024, 11:00-12:00: Living Lab Lecture #27: Beyond the surface: leveraging NLP to map global natural hazard impacts (online)
30.07.2024, 10:00-15:00: JupyterHub on HPC (ScaDS.AI online training)
22.08.2024 15:00-16:30: Graph Algorithms on Emerging Tile-Centric Accelerators (hybrid)
27.08.2024 10:00-11:00 am: A hitchhiker's guide to Ontology (APB-1004)
Softphone consultation hour (online):
July/August 2024: 01.07., 08.07., 15.07., 29.07., 05.08., 12.08., each 10:00-11:00 a.m.
Open Q&A session for users of the NHR@TUD computing clusters (online):
July/August 2024: 08.07., 22.07., 05.08., 19.08., 13:30-14:30 each time

http://www.tu-dresden.de/zih/veranstaltungen