Vonheiden, Björn (2021) Scalability Benchmarking of Kafka Streams Deployment Options. (Student research project), Kiel University, Kiel, 10 pp.

Preview

Text Bjoern-Vonheiden_Scalability-Benchmarking-of-Kafka-Streams-Deployment-Options.pdf - Published Version Download (297kB) | Preview

Abstract

Kubernetes, Kafka, and Kafka Streams are commonly used together in software systems. The Theodolite method allows to benchmark the scalability of such systems. Kafka Streams provides many configuration options. Thus, it is difficult to set the configuration options in the best possible way. Previous research with Theodolite, however, considers only few configuration options for the scalability. Therefore, we explore further configuration options for Kafka and Kafka Streams in this paper. On one hand we apply the topology optimization option, on the other hand we focus on high availability and resiliency options. We define different parameters for these configuration options and execute the Theodolite benchmark with them. Ultimately, we present and discuss the results of the benchmark executions and identify the influence of the parameters regarding the scalability of Kafka Streams applications.

Actions (login required)

View Item