Seminar_Distributed_Graph

Hosts

Fang Wei-Kleiner

Distributed Graph Processing Systems

Many of parallel graph processing systems have be developed to scale-out to a high number of computing nodes to accommodate very large graphs and high computation costs. Although most of the systems are based on the distributed file system such as HDFS (Hadoop Distributed File System), the design philosophy and the performance are different. In this seminar, we will discuss about these systems in details in an analytical manner. The systems are listed as follows (but not limited to): Hadoop/HaLoop, Pregel/Giraph, GraphLab, Spark/GraphX, Giraph++, Blogel, Vertica. Each member is supposed to select one system and find the related papers/tutorials for it.

Application Procedure

• We kindly ask prospective participants to apply via HISinOne under the name "Masterseminar am Lehrstuhl Datenbanken und Informationssysteme" for participating in this seminar.
• Please note that the number of participants is limited to 8.

Administrative Stuff

• The seminar is organized as a block seminar, consisting of a kick-off event and one or two presentation sessions with mandatory participation.
• Kick-off event:Wed. April 24th 2019, 16-18, Room 00-034, Geb 051
• The aim of the kick-off event is to provide all participants with basic knowledge about the seminar topics and organizational issues.
• The slides of the kick-off meeting can be found here
• Report templates: Latex / Word
• Additional information:
• ECTS: 4 cr. (equivalent to 2 SWS)
• Language: English/German