Jun 27, 2017





edit SideBar



FTA Toolbox

We have implemented a FTA toolbox to facilitate the comparative analysis of failure traces (see Figure). The toolbox is implemented in Matlab, and uses several open-source Matlab packages, such as the Mysql and DataTable packages.

The toolbox takes as input four functions for initializing, querying, processing, and finalizing the data analysis. The initialization and query stages allow one to extract the necessary data from traces located in the MySQL database into Matlab in-memory data structures. By contrast to loading entire data sets into memory from large files, this method allows one to extract into memory only the data that is needed for processing.

Initialization and querying is separated from processing to allow expensive initialization queries to be conducted only once, after which any amount of processing can be done. Also this separation allows the same initialization and queries to be used for many different processing functions. This facilitates code reuse.

The results of initializing and querying are then passed to the processing function. This function is run across each of those results. The processing output is then fed into the finalize function, which can produce tables automatically in latex, HTML, text, and wiki formats using the DataTable module


The toolbox is based on Matlab, and has been used with version (R2008a).

To connect to the database, users must first install a the matlab-mysql interface.

To print to tables in different formats, users should install the DataTable package.

You can check out the latest versions of the toolbox from the svn repository with the following command:

svn checkout svn://scm.gforge.inria.fr/svn/fta/toolbox

Matlab Scripts for tabbed files

Some useful Matlab scripts which is for event_trace.tab files (Matlab Version: (R2015a)).

  • Stats.m: Generate the all statisitcs parameters for event_trace.tab such as mean, variance, CV,...
  • plot_cdf.m: Plot CDF graph for availability and unavailability of trace in event_trace.tab.
  • fit_cdf.m: Fit various statistical models for CDF of availability and unavailability of trace in event_trace.tab.
  • qqplot2.m: Plot quantile-quantile graph. It is needed by "fit_cdf.m".


Some useful processing scripts are provided here.

Page Actions

Recent Changes

Group & Page

Back Links