README

This repository contains a project related to the paper How Bad Can a Bug Get? An Empirical Analysis of Software Failures in the OpenStack Cloud Computing Platform accepted for publication at the ESEC-FSE 2019 conference. The project includes tools to repeat the fault-injection experiments presented in the paper.

Please, cite the following paper if you use the tools for your research:

@inproceedings{cotroneo2019bad,
  title={How bad can a bug get? an empirical analysis of software failures in the OpenStack cloud computing platform},
  author={Cotroneo, Domenico and De Simone, Luigi and Liguori, Pietro and Natella, Roberto and Bidokhti, Nematollah},
  booktitle={Proceedings of the 2019 27th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering},
  pages={200--211},
  year={2019}
}

Please, note that this artifact does not include a fault injection tool since we transferred the ownership to our industry partners. Therefore, the artifact includes pre-injected source-code files. Before every fault injection test, an original source-code file is replaced with a pre-injected file, and it is restored after the test.

Project Organization

The diagram below provides the organization of the project:

|-- INSTALL.md
|-- LICENSE
|-- README.md
|-- data
|   -- Cinder.csv
|   -- Neutron.csv
|   -- Nova.csv
|-- results
|-- src
|   -- install
|       -- check_openstack.py
|       -- install_openstack.sh
|       -- prepare_packstack_config_file.sh
|       -- packstack_configuration_template.txt
|   -- run_ff.sh
|   -- run_test.sh
|   -- statistic.sh
|   -- tests
|       -- Cinder
|       -- Neutron
|       -- Nova
|   -- workload
|       -- cleanup.sh
|       -- restart_system.sh
|       -- start_workload.sh

The data directory contains the CSV files related to each OpenStack sub-system (i.e., Nova.csv, Cinder.csv, Neutron.csv). Each CSV file describes the all the information about each test case. Specifically, the CSV files have the following fields:

• Test_id: It is the name of the folder that contains the specific test, where id is an increasing number used to differentiate the tests;
• Fault_Type: It is the type of the injected fault from the list described in the paper;
• Target_Component: It is the name of the source code file to be mutated;
• Target_Class: It is the name of the class which contains the mutated statement;
• Target_Function: It is the name of the function which contains the mutated statement;
• Fault_Location: It is the target statement;
• Line: It is the line of the Fault_Location in the Target_Component.

The results directory will contain data generated during analysis.

The src directory contains all of the code written for the project, including the scripts needed for the installation of OpenStack (src/install sub-directory; see INSTALL.md file), the scripts used to execute the workload (src/workload sub-directory) and for analyzing the experiments.

The provided fault injection test suite is under src/tests directory. The test directory contains every test case sub-directory for each OpenStack sub-system (Nova, Cinder, Neutron) targeted during the experimentation provided in the paper. Every test case sub-directory contains the following files:

• fip_info.data: It stores information about the fault injected for the specific test;
• mutated_file: It contains the Target_Component after the mutation.