7. Treffen des Arbeitskreises MSDO

Programm

- Tag 1, 09.03.2020 (Mo)

• 13:00: Eintreffen
• 13:30: Willkommen, Organisatorisches
• 13:45: IVU.plan - ein Gigant wird Modulith (Andreas Noack, IVU)
• 14:30: Kaffeepause
• 15:00: From Domain-Driven Design to Microservice APIs of Quality and Style (Olaf Zimmermann, HSR Hochschule für Technik Rapperswil) [Folien]
• 15:45: Architectural Inter-Microservice Integration - An Overview (Georg Schwarz, Friedrich-Alexander University Erlangen-Nürnberg) [Folien]
• 16:30: Microservices und Enterprise-Software -- Wie passt das zusammen? (Holger Knoche, Universität Kiel) [Folien]
• ca. 17:30: Ende des offiziellen Teils des ersten Tages
   
• 19:00: Gemeinsames Abendessen (self-paid)
  Im Doldenmädel (Speisekarte, Website: https://doldenmaedel-berlin.de/)

- Tag 2, 10.03.2020 (Di)

• 09:00: Willkommen zu Tag 2
• 09:15: Blockchain and Services – Exploring the Links (Ingo Weber, Technische Universität Berlin). [Folien, Paper]
• 10:00: Zalando's Tech Journey (Rodrigue Schäfer, Zalando)
• 10:45: Kaffeepause
• 11:30: Experience in establishing a DevOps culture (Kai Sachs, Careem/Uber)
• 12:15: Chaos Experimentation based on Established Risk Analysis Methods: Experiences from an Industrial Case Study (Dominik Kesim, Uni Stuttgart) [Folien]
• 12:45: Abschluss
• 13:00: Ende des offiziellen Teils des zweiten Tages

Vortragsdetails:
 

• IVU.plan - ein Gigant wird Modulith
  Andreas Noack, IVU
  
  IVU.plan unterstützt hunderte Kunden weltweit bei der Planung und Disposition von öffentlichem Personenverkehr. Bei mehr als 10 Millionen Zeilen Quelltext und über 20 Jahren Entwicklungszeit ist Modularisierung überlebenswichtig. Der Vortrag stellt Prinzipien, Techniken und Herausforderungen auf dem Weg zum Modulithen vor.

• From Domain-Driven Design to Microservice APIs of Quality and Style
  Prof. Dr. Olaf Zimmermann, Institute for Software, HSR Hochschule für Technik Rapperswil
  
  Service designers, system integrators, and enterprise architects apply strategic Domain-driven Design (DDD) patterns such as Bounded Context and Context Map to identify service candidates. However, architectural design advice regarding service granularities, message representation structures, and API endpoint evolution is still missing. In this talk, we present a pattern language, two Domain-Specific Languages (DSLs) and supporting tools to fill this gap.

• Microservices und Enterprise-Software -- Wie passt das zusammen?
  Holger Knoche, Universität Kiel
  
  Viele bekannte Microservice-Anwendungen stellen verhältnismäßig einfache Prozesse für Endanwender zur Verfügung, wie beispielsweise Webshops oder Streamingdienste. Für derartige Anwendungen steht insbesondere die Elastizität bzw. Skalierbarkeit der Anwendung im Vordergrund, um der teils starken und schwer vorhersehbaren Dynamik des Nutzerverhaltens angemessen begegnen zu können. Im Gegensatz dazu bildet Enterprise-Software häufig komplizierte Prozesse ab und wird von einer bekannten und stabilen Anzahl professioneller Anwender genutzt. Entsprechend ist Skalierbarkeit für diese Anwendungen oftmals weniger relevant; statt dessen stehen Wartbarkeit und Evolvierbarkeit besonders im Vordergrund. Zudem bestehen auch teils andere Ansprüche an die Datenkonsistenz und Steuerung, beispielsweise durch Workflows.
  
  Im Rahmen dieses Vortrags sollen insbesondere folgende Fragen diskutiert werden:
  - Microservices und Enterprise-Software -- passt das überhaupt zusammen?
  - Welche Vorteile können Microservices für Enterprise-Software bieten?
  - Wo bestehen Schwierigkeiten, Microservices und Enterprise-Software zusammenzubringen?
  - Wie könnte eine grobe Referenzarchitektur für Microservice-basierte Enterprise-Software aussehen?

• Blockchain and Services – Exploring the Links
  Ingo Weber, TU Berlin
  
  Blockchain is a novel distributed ledger technology which has attracted a wide range of interests for building the next generation of applications. The broad range of blockchain applications is made possible by so-called smart contracts, which transform a blockchain system into a general compute platform. As a new paradigm and technology platform in the integration space, this bears the question to which degree there is a connection to services, and in how far lessons learned on services can be applied to Blockchain. In the talk, I explore four different facets of this topic, as follows. First, application-level service interfaces for interaction with blockchain-based applications enable easy integration with existing infrastructure. Second, service composition can be achieved through smart contracts, and enable different approaches to orchestrations and choreographies. Third, Blockchain-aaS offerings cover infrastructure operation, but the idea presents further possibilities. And finally, some principles for designing SOA services and microservices can be applied to smart contract design.

• Zalando's Tech Journey
  Rodrigue Schäfer, Zalando
  
  Zalando's tech journey from humble beginnings to microservices in the cloud. How business objectives drive tech decisions.

• Experience in establishing a DevOps culture
  Kai Sachs, Careem/Uber

• Architectural Inter-Microservice Integration - An Overview
  Georg Schwarz, Friedrich-Alexander University Erlangen-Nürnberg
  
  Microservices build a certain kind of distributed architecture with each microservice focusing on a distinct domain or bounded context of the whole application. In order to implement cross-cutting features the cooperation of multiple microservices is inevitable. Since the microservice-based architectural style is originally driven by practices that have proven to work well in industry, there are also opinions on how to do this cooperation between services over network. By analyzing popular microservice literature we give an overview on which architectural levels inter-microservice integration is possible. Furthermore, we discuss the potential consequences of this crucial architectural design decision.

• Chaos Experimentation based on Established Risk Anal-ysis Methods: Experiences from an Industrial Case Study
  Dominik Kesim, Universität Stuttgart
  
  Modern distributed systems more and more frequently adapt the microservice archi-tectural style to design cloud-based systems. Assessing the resilience of such sys-tems can be done by leveraging fault injections on the application- and network-level also known as resilience benchmarking. Chaos engineering is a new, yet, evolving discipline that forces a change in the perspective of how systems are developed with respect to their resilience. Whereas fault injection and testing approaches are said to be binary, i.e., an assertion is made about the system that is then verified to be true or false, chaos engineering is claimed to be exploratory. The key idea is to apply empirical experimentation in order to learn how a system behaves under turbulent conditions by intentionally injecting failures. In this talk, I will summarize the methods, results, and lessons learned from the case study.