PBlaman: Performance Blame Analysis based on Palladio Contracts
Paper written by Frank
Brüseke, Henning Wachsmuth, Steffen
Becker, Gregor
Engels
Case studies
The paper that this website refers to includes two case studies:- Please visit the website covering our previous work to read the details of our first case study dealing with the Common Component Modeling Example (CoCoME).
- The details of the second case study, which deals with an information extraction application implemented with Apache UIMA, are covered here.
- Tools and Data -
Prerequisites
- For the PCM-contracts and their simulation, please install Palladio
3.3: http://sdqweb.ipd.kit.edu/wiki/PCM_3.3
(not necessary for the data comparison part of the blame analysis) - For the data comparison of blame analysis
- please install R (tested with R 2.12.2): http://www.r-project.org/ or http://ftp5.gwdg.de/pub/misc/cran/
- put the R "bin"-directory on your PATH environment variable
- please install Python 2.7: http://www.python.org/download/releases/2.7.3/
- please install Java (tested with Java 7): http://www.java.com (only for conversion of test data series)
Step-By-Step-Guide
In this section we give a step by step guide, which describes the decision support steps of the PBlaman approach as they need to be performed by the system architect. There are several short cuts as the links above include a lot of intermediate results.
Preparation
Create a dedicated folder for the data, scripts, and derived
visualizations. This dedicated folder will be referred to as
"<dataPath>".
Guide
- Part 1: Palladio simulation
- Part 2: Test result analysis
& comparison
- short cut to "Part 2":
unpack the PCM-Data series in your <dataPath>
All Downloads
Licenses and Copyright
- The scripts and models on this website are made available under the regulations of the GNU Lesser General Public License LGPL
- The data sets, the resulting diagrams and tables, and also this website are made available under the regulations
of the Creative Commons Attribution-ShareAlike 3.0 Unported License
Data
- Test data series:
- PCM (aka Palladio) data series: [Zip]
- PCM to Implementation mapping: [File]
- Conversion factor configuration file: [File]
Models
- Palladio-Models for the Palladio 3.3 workbench: [Zip]
Scripts
- Data preparation:
- Test data series conversion application (Java): [Zip]
- Evaluation scripts:
- All-in-one: [Zip]
Output
- The performance report: analysis1.RR.pdf
- The extended performance report: analysis1.R.pdf
- The flame graph: analysis1.flame.csv2.svg
