Thursday, 14 December 2017

Session 1: Workshop Presentations

Modeling Variations in Load Intensity over Time

Authors:

Jóakim v. Kistowski (Karlsruhe Institute of Technology)
Nikolas Herbst (Karlsruhe Institute of Technology)
Samuel Kounev (Karlsruhe Institute of Technology)

Abstract:

Today’s software systems are expected to deliver reliable performance under highly variable load intensities while at the same time making efficient use of dynamically allocated resources. Conventional benchmarking frameworks provide limited support for emulating such highly variable and dynamic load profiles and workload scenarios. Industrial benchmarks typically use workloads with constant or stepwise increasing load intensity, or they simply replay recorded workload traces. Based on this observation, we identify the need for means allowing flexible definition of load profiles and address this by introducing two meta-models at different abstraction levels. At the lower abstraction level, the Descartes Load Intensity Meta-Model (DLIM) offers a structured and accessible way of describing the load intensity over time by editing and combining mathematical functions. The High-Level Descartes Load Intensity Meta-Model (HLDLIM) allows the description of load variations using few defined parameters that characterize the seasonal patterns, trends, bursts and noise parts. We demonstrate that both meta-models are capable of capturing real-world load profiles with acceptable accuracy through comparison with a real life trace.

DOI: 10.1145/2577036.2577037

Full text: PDF

[#][]

Using Performance Models to Support Load Testing in a Large SOA Environment

Authors:

Christian Vögele (fortiss GmbH)
Andreas Brunnert (fortiss GmbH)
Alexandru Danciu (fortiss GmbH)
Daniel Tertilt (fortiss GmbH)
Helmut Krcmar (Technische Universität München)

Abstract:

Load testing in large service-oriented architecture (SOA) environments is especially challenging when services are under the control of different teams. It gets even more difficult if services need to be scaled before a load test starts. It is thus important to estimate workloads for services involved in a load test. Service workloads can be specified by the amount of service operation invocations distributed over time. We propose the use of performance models to derive this information for SOA-based applications before executing load tests. In a first step, we use these models to select usage scenarios. Afterwards, these models are transformed in a way that each scenario can be simulated separately from each other. These simulations can predict service workloads for selected usage scenarios and different user counts.

DOI: 10.1145/2577036.2577038

Full text: PDF

[#][]