Wednesday, 16 August 2017

Storage Performance Analyzer (SPA)

Description

The Storage Performance Analyzer (SPA) is a software package containing the functionality for the systematic measurement, analysis and regression modeling specifically tailored for storage systems. SPA consists of a benchmark harness that coordinates and controls the execution of the included I/O benchmarks (FFSB and Filebench) and a tailored analysis library used to process and evaluate the collected measurements.

The benchmark harness supports the execution of benchmarks. Using a specification of the benchmark parameter space, which should be run on the remote machine, the software ensures that each of these configurations is executed automatically in a deterministic order. Furthermore, the benchmark harness supports synchronous measurements on co-located virtual machines in virtualized environments. The software ensures that the benchmark results are stored persistently in an SQLite database to provide easy access for further analysis.

The analysis library provides the functionality for the analysis of the measurement results obtained with the benchmark harness. By providing a drop-in solution for the widely accepted statistical framework R, the power of R is combined with the specialized functionality of SPA. The analysis functions help with the derivation and analysis of regression models from the benchmark results. It also contains functionality to optimize many regression techniques to maximize the modeling power in different scenarios. The R environment can be exploited to further analyze the measurements and to generate illustrations of the analysis results.

Requirements
  • Java 6 or later
  • Apache Ant for compilation
  • Controller Machine: Windows, Unix, or MacOS
  • Target Machine: POSIX compatible OS
  • Data post-processing: R Project, SQLite 3
Contributor(s)
  • Qais Noorshams
  • Axel Busch
  • Dominik Bruhn
  • Samuel Kounev
  • Ralf Reussner
Maintainer(s)
  • Qais Noorshams, noorsh(at)ira.uka.de

    Developed at:
    Karlsruhe Institute of Technology (KIT), Chair for Software Design and Quality
    Am Fasanengarten 5, 76131 Karlsruhe, Germany

  • Samuel Kounev, samuel.kounev(at)uni-wuerzburg.de

    University of Wuerzburg, Chair of Computer Science II
    Am Hubland, 97074 Würzburg, Germany
Version
  • 1.0
Download
License
  • GNU General Public License (GPL), Version 2.0 or later
Website
Publications/ Projects using the tool
  • Qais Noorshams, Roland Reeb, Andreas Rentschler, Samuel Kounev, and Ralf Reussner. Enriching Software Architecture Models with Statistical Models for Performance Prediction in Modern Storage Environments, in: Proceedings of the 17th International ACM Sigsoft Symposium on Component-Based Software Engineering, CBSE '14, Lille, France, 2014. ACM, New York, NY, USA. 2014.
  • Qais Noorshams, Axel Busch, Andreas Rentschler, Dominik Bruhn, Samuel Kounev, Petr Tůma, and Ralf Reussner. Automated Modeling of I/O Performance and Interference Effects in Virtualized Storage Systems, in: 34th IEEE International Conference on Distributed Computing Systems Workshops (ICDCS 2014 Workshops). 4th International Workshop on Data Center Performance, DCPerf '14, Madrid, Spain, 2014. IEEE Computer Society.
  • Qais Noorshams, Kiana Rostami, Samuel Kounev, Petr Tůma, and Ralf Reussner. I/O Performance Modeling of Virtualized Storage Systems, in: Proceedings of the IEEE 21st International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems, MASCOTS '13, San Francisco, USA, 2013, pages 121-130. IEEE Computer Society.
  • Qais Noorshams, Dominik Bruhn, Samuel Kounev, and Ralf Reussner. Predictive Performance Modeling of Virtualized Storage Systems using Optimized Statistical Regression Techniques, in: Proceedings of the ACM/SPEC International Conference on Performance Engineering, ICPE '13, Prague, Czech Republic, 2013, pages 283-294. ACM, New York, NY, USA. 2013.