Datenschutzerklärung|Data Privacy
Impressum

15.08.2019
L. Friedel

Philipp Grulich and Jonas Traub present the Scotty Window Processor at Code.Talks 2019 in Hamburg

This year's Code.Talks conference will take place on the 24. and 25. October in the center of Hamburg. With more than 1500 participants, 8 Stages, and more than 120 Talks in two days, the event is the largest developer conference in Germany and one of the larges developer conferences in Europe. Code.Talks puts a strong focus on technical aspects of software engineering and innovative software technology.
Jonas Traub and Philipp Grulich will present how they designed an developed the Scotty Window Processor. Scotty enables efficient window aggregation for diverse stream processing system and is the result of a research project at the TU Berlin DIMA group and the BBDC. Window aggregation is a core operation in data stream processing. Stream Processing Systems, like Flink or Storm, implement general aggregation techniques which perform poorly under specific workloads (e.g. Sliding Windows). Scotty is new highly-efficient window operator, which exploits specific workload properties such as the type of aggregation functions (e.g., invertible, associative), window types (e.g., sliding, sessions), windowing measures (e.g., time- or count-based), and stream (dis)order. This allows Scotty to outperform systems like Flink by up to one order of magnitude. Scotty and its connectors are available as open-source and contributions are highly welcome.

Open Source Repository: scotty-window-processor
[Scotty Paper at ICDE 2018: http://www.user.tu-berlin.de/powibol/assets/publications/traub-scotty-icde-2018.pdf|Scotty Paper at ICDE 2018]. More on experiments with out-of-order stream processing from DEBS
2019:[ http://www.user.tu-berlin.de/powibol/assets/publications/Grulich-Reproducible-out-of-order-streams-paper.pdf|Grulich-Reproducible-out-of-order-streams-paper]


Event:
Code.Talks
24. - 25. Oktober 2019
CinemaxX Dammtor
Dammtordamm 1
20354 Hamburg