Navigation Zur Modulseite
Anzeigesprache

In-Memory Databases On Modern Hardware
Hauptspeicherdatenbanken auf moderner Hardware

6 LP

Deutsch

#40804 / #1

Seit SS 2017

Fakultät IV

EN 7

Institut für Softwaretechnik und Theoretische Informatik

34351500 FG Datenbanksysteme und Informationsmanagement

Markl, Volker

Breß, Sebastian

sebastian.bress@dfki.de

Keine Angabe

POS-Nummer PORD-Nummer Modultitel
2346700 38129 In-Memory Databases On Modern Hardware

Lernergebnisse

Traditionally, database systems managed databases that were primarily stored on secondary storage and only a small part of the data could fit in main memory. Therefore, disk IO was the dominating cost factor. Nowadays, it is possible to equip servers with several terabytes of main memory, which allows us to keep databases in main memory to avoid the disk IO bottleneck. Therefore, the performance of database systems became limited by memory access and processing power. This course will teach students the fundamentals of efficient data processing in main-memory database systems using techniques optimized for main memory (e.g., column stores and query compilation) and modern processor capabilities (e.g., SIMD-based database algorithms, GPU co-processing).

Lehrinhalte

The course is split into two parts, each covering roughly one half of the semester. During the first part, the students learn the fundamentals of cache-efficient storage and processing models. This includes columnar storage and query processing, compression, vector-at-a-time processing, query compilation and transaction processing. In the second part, students learn the basics of parallel data processing on modern CPUs and co-processors (e.g., GPUs) for typical database operators, including optimizations such as SIMD and NUMA-awareness. The course consists of a lecture and theoretical, written exercises.

Modulbestandteile

Pflichtgruppe:

Die folgenden Veranstaltungen sind für das Modul obligatorisch:

Lehrveranstaltungen Art Nummer Turnus Sprache SWS
In-Memory Databases On Modern Hardware IV 3435 L 9195 SS Keine Angabe 4

Arbeitsaufwand und Leistungspunkte

In-Memory Databases On Modern Hardware (IV):

Aufwandbeschreibung Multiplikator Stunden Gesamt
Präsenzzeit 15.0 4.0h 60.0h
60.0h (~2 LP)

Lehrveranstaltungsunabhängiger Aufwand:

Aufwandbeschreibung Multiplikator Stunden Gesamt
Exam preparation 1.0 30.0h 30.0h
Graded problem sheets 1.0 60.0h 60.0h
Preparation/Consolidation IDB 15.0 2.0h 30.0h
120.0h (~4 LP)
Der Aufwand des Moduls summiert sich zu 180.0 Stunden. Damit umfasst das Modul 6 Leistungspunkte.

Beschreibung der Lehr- und Lernformen

Lectures are accompanied by individual exercises to practically rehearse the theory taught in the lectures. The course will be given in English.

Voraussetzungen für die Teilnahme / Prüfung

Wünschenswerte Voraussetzungen für die Teilnahme an den Lehrveranstaltungen:

This course is an advanced course for master students with focus on database systems and information management. In contrast to the introduction of database systems (MPGI5/DBS/Informationssystemse&Datenanalyse), which looks at database systems from an application programmers point of view, this class focuses on the internals of in-memory database systems and various optimizations for efficient query processing on modern hardware. It is useful (but not required) to attend the course Database Technology (DBT) prior to this course. To participate, students are required to have successfully completed a Bachelor in computer science with a focus on database systems (participation in the Datenbankpraktikum, Datenbankprojekt). Knowledge of data modeling, relational algebra, and SQL as well as a very good command of Java, or possibly C/C++/C#, programming is required to participate in the course. Due to capacity reasons, the class is limited to at most 60 participants.

Verpflichtende Voraussetzungen für die Modulprüfungsanmeldung:

Keine Angabe

Abschluss des Moduls

Benotung:

benotet

Prüfungsform:

Portfolioprüfung

Sprache:

Deutsch

Art der Portfolioprüfung

100 Punkte insgesamt

Prüfungselemente

Name Punkte Kategorie Dauer/Umfang
(Deliverable assessment) Assessment of 4 homework exercises with 5 pts 20 schriftlich 4x15h = 60h
(Examination) End term test 40 schriftlich 60 min.
(Examination) Mid term test 40 schriftlich 60 min.

Notenschlüssel

1.01.31.72.02.32.73.03.33.74.0
95.090.085.080.075.070.065.060.055.050.0

Prüfungsbeschreibung (Abschluss des Moduls)

Keine Angabe

Dauer des Moduls

Dieses Modul kann in einem Semester abgeschlossen werden.

Maximale teilnehmende Personen

Die maximale Teilnehmerzahl beträgt 60.

Anmeldeformalitäten

Students are required to register via the DIMA course registration tool before the start of the first lecture (http://www.dima.tu-berlin.de/). Within the first six weeks after commencement of the lecture, students will have to register for the course at QISPOS (university examination protocol tool) and ISIS (course organization tool) in addition to the registration at the DIMA course registration tool.

Literaturhinweise, Skripte

Skript in Papierform

Verfügbarkeit:  nicht verfügbar

Skript in elektronischer Form

Verfügbarkeit:  nicht verfügbar

Literatur

Empfohlene Literatur
Alfons Kemper, André Eickler Datenbanksysteme. Eine Einführung.10., aktualisierte und erweiterte Auflage, Oldenbourg Verlag, 2015.
Daniel Abadi, Peter A. Boncz, Stavros Harizopoulos, Stratos Idreos, Samuel Madden: The Design and Implementation of Modern Column-Oriented Database Systems. Foundations and Trends in Databases 5(3): 197-280 (2013)
Hasso Plattner. 2014. A Course in In-Memory Data Management: The Inner Mechanics of In-Memory Databases. Second Edition. Springer Publishing Company, Incorporated.
John L. Hennessy, and David A. Patterson. Computer architecture: a quantitative approach. Elsevier, 2012.

Zugeordnete Studiengänge

Dieses Modul wird auf folgenden Modullisten verwendet:

Sonstiges

Keine Angabe