9-12 May 2017
Europe/Rome timezone

VIP, a simplified access to high-level resources

Not scheduled
15m

Speakers

Axel Bonnet (CNRS) Mr Pascal Girard (CNRS)

Description

VIP (Virtual Imaging Platform) is a web portal for medical images and data analysis. It uses the resources of the biomed Virtual Organisation of the EGI e-infrastructure to execute applications directly from the web (Application as a service). As of April 2017, VIP has more than 1000 registered users and about 20 applications open to all of them. Furthermore, VIP was in 2016 the executing platform of 2 challenges of the prestigious MICCAI conference. Via a user friendly GUI, VIP allows to easily launch applications requiring important hardware resources, on several datasets at the same time. VIP can split an execution in multiple jobs (data parallelism) in a transparent manner for the end user. The user can monitor and manage on the web the executions he has launched. The pipeline creator can easily allow other users to execute his pipeline, thus making it shareable. This avoids installation and portability issues. New pipelines are regularly added into VIP thanks to a simple application import mechanism, called Boutiques, supported by the platform. Boutiques is an application repository that allows automatic import and exchange of applications in data analysis platforms. It relies on Linux containers (Docker, Singularity) to solve the problem of application installation in a lightweight manner and it uses a versatile JSON format to describe command line tools. An ongoing effort is the implementation of a REST API, based on the CARMIN specification that will allow VIP to be used transparently from external tools. Several projects are already interested in this feature,in order, for instance, to control complex applications executing on EGI directly from the user environment. This will multiply the use cases of VIP and expose its applications to many new people. This poster will emphasize the above-mentioned features of VIP which improve the user experience.

Primary authors

Axel Bonnet (CNRS) Mr Frederic Cervenansky (CNRS) Mr Pascal Girard (CNRS) Sorina POP (CNRS) Tristan Glatard (CNRS)

Presentation Materials