Ozyegin University, Istanbul (TURKEY)
University of Twente, Enschede (THE NETHERLANDS)
European Space Agency, Noordwijk (THE NETHERLANDS)
CADP (Construction and Analysis of Distributed Processes)
The availability of the system, that is, the percentage of time
the system is up and running, heavily depends on the chosen software
decomposition alternative, i.e., the way in which the software
modules are grouped into recoverable units (RUs). The FLORA
framework (see http://cadp.inria.fr/software/09-b-flora.html)
supports the implementation of local recovery for a particular
software decomposition. Despite the support of this framework,
the implementation of local recovery for a decomposition alternative
is still a time-consuming and a non-trivial task. Hence, it is
important to follow a systematic method to compare various
alternatives and only implement the best one. A software architecture
decomposition can be optimized for local recovery based on quality
estimations for decomposition alternatives. However, optimization
needs input from a systematic and sound method that provides
availability estimations regarding decomposition alternatives at
This work presents such a method, which takes as input a decomposition of the software architecture (the set of modules, the failure and repair rates for each module, and the grouping of modules into RUs). A new tool named Availability Analyzer provides an availability estimation for a given decomposition alternative based on analytical models. Once performance and availability estimations are known regarding each decomposition alternative, the Optimizer tool can select the best alternative. The selected alternative is then implemented with the FLORA framework. Availability Analyzer is implemented in Java and it makes use of the CADP toolset. It takes as input a decomposition alternative and a set of predefined MIOA specifications, which formally define the failure and recovery behavior of system modules. First, Availability Analyzer generates a set of I/O-IMC models corresponding to the given decomposition alternative. It also generates a composition script in SVL, which specifies how these models must be combined. Then, the SVL script is executed with the CADP toolset to obtain a single CTMC model by composing all the I/O-IMC models. Finally, the CTMC model is used for providing an availability estimation.
This modeling and analysis approach is illustrated on a real-life software system, namely the MPlayer open-source media player. Four different decomposition alternatives have been investigated and the availability predicted by the analytical models was compared to the availability measurements obtained from the actual implementations. It was observed that the analytical results closely match the measured availabilities.
The availability analysis methodology proposed follows a modular and
compositional approach, which is well-established for software design
but less often used for analytical purposes. In the proposed approach,
separate analytical models are generated for different architectural
elements. These models are composed together based on how the
architectural elements are composed with each other. Different
architectural configurations can be evaluated by just using
(generating) a different composition script working on the same
set of models. If failure behavior of a module or the recovery
strategy changes, such changes should only be reflected to the
corresponding analytical model that is defined separately from
the other models. A CTMC model for the overall system is generated
in multiple steps, thanks to the model composition and reduction
features of CADP.
Hasan Sozer, Marielle Stoelinga, Hichem Boudali, and Mehmet Aksit.
"Availability Analysis of Software Architecture Decomposition
Alternatives for Local Recovery".
Software Quality Journal, pp. 1-27, May 2016.
Available on-line at: http://link.springer.com/article/10.1007/s11219-016-9315-9
or from our FTP site in PDF or PostScript
Dept. of Computer Science (Zilverling)
Formal Methods and Tools group
P.O. Box 217
7500 AE Enschede
Tel: +31 53 489 3773
|Further remarks:||This tool, amongst others, is described on the CADP Web Page: http://cadp.inria.fr|