Test derivation for the software defined networking platforms: novel fault models and test completeness N. V. Yevtushenko, I. Burdonov, A. Kossachev [et al.]
Material type:
ArticleSubject(s): программно-определяемые сети | сети передачи данных | виртуализация сетиGenre/Form: статьи в сборниках Online resources: Click here to access online
In:
Proceedings of IEEE East-West Design & Test Symposium (EWDTS'2018), Kazan, Russia, September 14-17, 2018 P. 358-363Abstract: Software Defined Networking (SDN) platforms are
used in many applications and thus, should be
thoroughly verified and tested before deployment;
there are a number of research papers proposing
various approaches for their verification. The next
step is to test whether all the items are proceeding
properly, i.e., to test the platform. As the requests can
be considered as sets of paths which have to be
properly implemented, the inputs of the System Under
Test (SUT) are appropriate graphs (and packets
traversing those graphs) while the output that can
also be considered as a set of paths with appropriate
parameters, is obtained via the appropriate platform
monitoring. Correspondingly, for testing, we use a socalled
proactive approach: check a submitted set of
paths after forwarding a packet with appropriate
parameters. We define several fault models (under
different testing assumptions) with respect to the
underlying resource connectivity graph (RNCT). Some
conditions for deriving a complete test suite as well as
the complexity upper bounds with respect to defined
fault models are also established.
Библиогр.: 7 назв.
Software Defined Networking (SDN) platforms are
used in many applications and thus, should be
thoroughly verified and tested before deployment;
there are a number of research papers proposing
various approaches for their verification. The next
step is to test whether all the items are proceeding
properly, i.e., to test the platform. As the requests can
be considered as sets of paths which have to be
properly implemented, the inputs of the System Under
Test (SUT) are appropriate graphs (and packets
traversing those graphs) while the output that can
also be considered as a set of paths with appropriate
parameters, is obtained via the appropriate platform
monitoring. Correspondingly, for testing, we use a socalled
proactive approach: check a submitted set of
paths after forwarding a packet with appropriate
parameters. We define several fault models (under
different testing assumptions) with respect to the
underlying resource connectivity graph (RNCT). Some
conditions for deriving a complete test suite as well as
the complexity upper bounds with respect to defined
fault models are also established.
There are no comments on this title.