Tuesday, 12 December 2017

Session 3: Technical Papers

Automatic Extraction of Session-Based Workload Specifications for Architecture-Level Performance Models

Authors:

Christian Vögele (fortiss GmbH)
André van Hoorn (University of Stuttgart)
Helmut Krcmar (Technische Universität München)

Abstract:

Workload specifications are required in order to accurately evaluate performance properties of session-based application systems. These properties can be evaluated using measurement-based approaches such as load tests and model-based approaches, e.g., based on architecture-level performance models. Workload specifications for both approaches are created separately from each other which may result in different workload characteristics. To overcome this challenge, this paper extends our existing WESSBAS approach which defines a domain-specific language (WESSBAS-DSL) enabling the layered modeling and automatic extraction of workload specifications, as well as the transformation into load test scripts. In this paper, we extend WESSBAS by the capability of transforming WESSBAS-DSL instances into workload specifications of architecture-level performance models. The transformation demonstrates that the WESSBAS-DSL can be used as an intermediate language between system-specific workload specifications on the one side and the generation of required inputs for performance evaluation approaches on the other side. The evaluation using the standard industry benchmark SPECjEnterprise2010 shows that workload characteristics of the simulated workload match the measured workload with high accuracy.

DOI: 10.1145/2693182.2693183

Full text: PDF

[#][]

Model-based Performance Evaluation of Large-Scale Smart Metering Architectures

Authors:

Johannes Kroß (fortiss GmbH)
Andreas Brunnert (fortiss GmbH)
Christian Prehofer (fortiss GmbH)
Thomas A. Runkler (Siemens AG)
Helmut Krcmar (Technische Universität München)

Abstract:

Smart meter devices are used to monitor and control energy consumption and are interlinked with smart grids. Their growing use leads to an extensive amount of available data to be processed and causes smart grids to evolve to large-scale systems of systems. Guaranteeing appropriate scalability and performance characteristics is a tremendous challenge. In this paper, we focus on the provisioning of sufficient computing capacity to efficiently analyze the produced data in such a distributed system. For this purpose, we show the use of performance models to plan and simulate this distributed computation in smart grid systems. It demonstrates how different system architectures can be evaluated and required capacities can be estimated to cope with the occurring data volume. We analyze response times for time-critical tasks and assess the scalability of smart grid systems.

DOI: 10.1145/2693182.2693184

Full text: PDF

[#][]

High-Volume Performance Test Framework using Big Data

Authors:

Michael Yesudas (IBM Corporation)
Girish Menon S (IBM United Kingdom Limited)
Satheesh K Nair (IBM India Private Limited)

Abstract:

The inherent issues with handling large files and complex scenarios cause the data-driven approach [1] to be rarely used for performance tests. Volume and scalability testing of enterprise solutions typically requires custom-made test frameworks because of the complexity and uniqueness of data flow. The generation, transformation and transmission of large sets of data pose a unique challenge for testing a highly transactional back-end system like the IBM Sterling Order Management (OMS). This paper describes a test framework built on document-oriented NoSQL database, a design that helps validate the functionality and scalability of the solution simultaneously. This paper also describes various phases of planning, development, and testing of the OMS solution that was executed for a large retailer in Europe to test an extremely high online sales scenario. An out-of-the-box configuration of the OMS with the feature support for database sharding was used to drive scalability. The exercise was a success, and it is the world’s largest IBM Sterling Order Management benchmark in terms of sales order volume, to date.

DOI: 10.1145/2693182.2693185

Full text: PDF

[#][]