Logo des Repositoriums

Auflistung nach Autor:in "Rank, Johannes"

1 - 5 von 5
  • Konferenzbeitrag
    A Dynamic Resource Demand Analysis Approach for Stream Processing Systems
    (Softwaretechnik-Trends Band 40, Heft 3, 2020) Rank, Johannes; Hein, Andreas; Krcmar, Helmut
    Systems that provide real-time business insights based on live data, so-called Stream Processing Systems (SPS), have received much attention in recent years. In many areas such as stock markets or surveillance, it is essential to process data immediately and react accordingly. As the processing of real-time data is at the heart of SPS, their performance in terms of latency, throughput, and resource utilization constitutes a crucial role. Traditional performance and benchmarking approaches for SPS usually focus on the throughput and latency, trying to answer the question of which engine processes the incoming events fastest. However, neglecting the corresponding resource utilization provides only a limited and sometimes even misleading view on their actual performance. Depending on the use-case, an engine that achieves faster processing results at the cost of higher memory utilization is not always best suited, which can be shown based on the example of IoT edge computing devices with limited resources. For this reason, we developed a dynamic performance approach to analyze the resource demands of an SPS. The approach yields fine-grained performance metrics based on the individual processing steps of the SPS and without requiring any knowledge of the actual source code. More-over it takes the whole system (engine and streaming application) into account. Since, we do not rely on code instrumentation or language-specific profiling techniques but instead, use the dynamic tracing capabilities of the Linux kernel, we can support a broad range of different SPSs. We evaluate our approach by inspecting the CPU performance of Apache Flink while performing the Yahoo streaming benchmark.
  • Konferenzbeitrag
    Predicting Scaling Efficiency of Distributed Stream Processing Systems via Task Level Performance Simulation
    (Softwaretechnik-Trends Band 43, Heft 1, 2023) Rank, Johannes; Barnert, Maximilian; Hein, Andreas; Krcmar, Helmut
    Stream processing systems (SPS) are a special class of Big Data systems that firms employ in (near) real time business scenarios. They ensure low-latency processing through a high degree of parallelization and elasticity. However, firms often do not know which scaling direction: horizontally, vertically, or mixed, is the best strategy in terms of CPU performance to scale those systems. Especially in cloud deployments with a pay-per-use model and cluster sizes that can span dozens of cores and machines, firms would profit from more accurate measurement-based approaches. In this paper, we show how to predict the CPU consumption of Apache Flink for different scaling scenarios using the Palladio Component Model. Our approach models the individual streaming tasks that make up the application and parametrizes it with fine grained CPU metrics obtained by combining BPF pro filing and querying the CPU’s performance measurement unit. Through this “task-level model approach”, we can achieve highly accurate predictions, despite using a simple model and only requiring a few mea surements for parametrization. Our experiment also shows that we achieve more accurate results than an alternative approach based on regression analysis.
  • Konferenzbeitrag
    The Role of Performance in Streaming Analytics Projects: Expert Interviews on Current Challenges and Future Research Directions
    (Softwaretechnik-Trends Band 43, Heft 1, 2023) Rank, Johannes; Hein, Andreas; Krcmar, Helmut
    Stream processing systems (SPS) are becoming more frequent due to current trends such as Industry 4.0 or the Internet of Things. These systems’ performance is particularly important, as their timely processing is a crucial capability. At the same time, these systems are often combined with novel machine learning approaches (steaming analytics) that have high performance demands. This combination poses potential challenges for performance management. In this paper, we have conducted expert interviews in the industry to identify performance challenges in streaming analytics implementations and to derive future research directions to address them. Our analysis shows that while the experts had different opinions on the role of performance in project management, they agreed on five common challenges.
  • Konferenzbeitrag
    Towards Model-based Performance Predictions of SAP Enterprise Applications
    (Softwaretechnik-Trends Band 39, Heft 3, 2019) Streitz, Adrian; Barnert, Maximilian; Rank, Johannes; Kienegger, Harald; Krcmar, Helmut
    High-performing Enterprise Applications are the basis for efficient running business processes. In order to evaluate software performance, traditional methods refer to complex test scenarios following the development phase and neglect that problems are easier fixable when discovered early. This paper tackles the problem of late performance evaluations and presents a conceptual approach that enables response time predictions for SAP Enterprise Applications during the development phase. We introduce a performance model generator that transforms ABAP source code into Palladio Component Model instances by using Abstract Syntax Trees which allows to conduct early performance simulations. Our approach supports conditional and probabilistic control flows to improve prediction accuracy. Based on subsequent performance simulations, we predict response time of applications and their underlying processing systems.
  • Konferenzbeitrag
    Using OPEN.xtrace and Architecture-Level Models to Predict Workload Performance on In-Memory Database Systems
    (Softwaretechnik-Trends Band 39, Heft 4, 2019) Barnert, Maximilian; Streitz, Adrian; Rank, Johannes; Kienegger, Harald; Krcmar, Helmut
    In-Memory Database Systems (IMDB) come into operation on highly dynamic on-premise and cloud environments. Existing approaches use classical modeling notations such as queuing network models (QN) to reflect performance on IMDB. Changes to workload or hardware come along with a recreation of entire models. At the same time, new paradigms for IMDB increase parallelism within database workload, which intensifies the effort to create and parameterize models. To simplify and reduce the effort for researchers and practitioners to model workload performance on IMDB, we propose the use of architecture level performance models and present a model creation process, which transforms database traces of SAP HANA to the Palladio Component Model (PCM). We evaluate our approach based on experiments using analytical workload. We receive prediction errors for response time and throughput below 4 %.