Nur in mobiler Ansicht

Modul Ultra Large Scale Systems

(Ultra Large Scale Systems)

Modul:

143102 Ultra Large Scale Systems (Wahlpflichtmodul im Grundstudium), Schwerpunkte: Mobile Media and Networks, Software Technology und Engineering

Modulverantwortlicher:

Prof. Dr. Tobias Jordine

ECTS-min./max.:

5 / 5

Workload:

Siehe Beschreibung der zugehörigen LV(s)

Modulprüfung:

Formale Zulassungsvoraussetzungen:

siehe SPO

Kompetenzprofil

Lernergebnisse:

The course assumes, that the participants are part of Amazon's or Google's architecture team, responsible for scalability, performance, availability and operations of large scale distributed infrastructures. The technical aspects need to be seen in context of fast changing business requirements. We start with an analysis of existing ULS sites and identify patterns and architectures that allow extreme growth. Participants learn to analyse bottlenecks and critical paths and develop methodologies for modeling and calculating performance. They learn to partition system aspects into horizontally scalable parts. After an analysis phase, the course dives into core components of large scale distributed systems and the participants learn the mechanics of scheduling, storage, locking, caching and asynchronous queues. They also learn to apply those components in solutions for small and mid-size companies. Below components, a layer of algorithms forms the base of performance and availability. Distributed consensus with fault-tolerance, asynchronous I/O methods and concurrency and parallelism belong into this layer. Participants learn how to apply those algorithms in the construction of scalable components. Data form a special part of ULS architectures and participants will learn how to model data with respect to large numbers of concurrent updates. Monitoring and System management are essential for ULS and participants will learn about feedback control mechanisms. They understand the architecture needed for monitoring and changing ULS systems. At the lowest level, failure modes and time in distributed systems are the most important theoretical concepts that need to be mastered. Participants will learn about the core mechanisms of a distributed system After the course, participants are able to analyse scalability and performance problems in current systems. They are able to design new systems with some degree of scalability built into them right from the beginning. They understand the critical mapping of business requests to infrastructure. They understand critical components and how to integrate them into existing architectures. Besides the technical competencies students acquire the ability to discuss various options and alternatives within teams and with external professionals and develop an independent position based on a clear scientific unterstanding of the subject matter. In a permanently changing area this is a vital requirement for success. This includes the ability to question not only the technical foundation but also business requirements and goals.

Weitere Angaben einblenden

Voraussetzung für dieses Modul:

Dieses Modul ist Voraussetzung für:

beinhaltet folgende Lehrveranstaltung(-en):

EDV-Nr

Lehrveranstaltung

Art

SWS

ECTS

Prüfungsform

143102a

Ultra Large Scale Systems

* kennzeichnet Prüfungsvorleistungen