BEGIN:VCALENDAR VERSION:2.0 PRODID:icalendar-ruby CALSCALE:GREGORIAN METHOD:PUBLISH BEGIN:VTIMEZONE TZID:Europe/Paris BEGIN:DAYLIGHT DTSTART:20160327T030000 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3 TZNAME:CEST END:DAYLIGHT BEGIN:STANDARD DTSTART:20161030T020000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10 TZNAME:CET END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP;TZID=Europe/Paris:20160424T054755 UID:11224@assets1.agendadulibre.org DTSTART;TZID=Europe/Paris:20160517T093000 DTEND;TZID=Europe/Paris:20160517T190000 DESCRIPTION:Quatrième événement du [Printemps de l'innovation Open Sourc e][1]\, organisé par le GTLL de Systematic et l'Irill\, présidé par Rob erto Di Cosmo\n\n[Open Source pour le Cloud et les Conteneurs][2]\n\nProgr amme dirigé par **Frédéric Lepied** (Red Hat) et **Gilles Muller** (Inr ia / LIP6)\n\n* 10:00 - Everything you want to know about virtualization o n large multicores\n **Vivien Quéma** (IMAG)\, **Gaël Thomas (Telecom S ud-Paris)\, Pierre Sens** (LIP6)\, **Gauthier Voron** (LIP6)\n\n Cloud pr oviders are now proposing large virtual machines with several tens of cor es. Since such virtual machines often relies on complex hardware architec tures\, new challenges arise for hypervisors to achieve good performances. This talk will present some of these problems and why they happen.\n\n* 10:45 - Reclaiming memory to the right containers\n **Julien Sopena** (LI P6)\n\n* 11:30 - Towards microservices architecture to transcode videos in the large at low costs\n **Olivier Barais** et **David Bromberg** (IRISA ).\n\n The increasing popularity of videos over Internet\, combined with the wide heterogeneity of various kinds of end users’ devices\, imposes strong requirements on the underlying infrastructure and computing resourc es to meet the users expectations. In particular\, designing an adequate t ranscoding workflow in the cloud to stream videos at large scale is: (i) c ostly\, and (ii) complex. Lightweight containers technologies such as dock er are more and more adopted for deploying and managing micro-services arc hitectures \\[1\\]. Indeed\, they hide a part of the complexity resulting from the language heterogeneity and the large choice of existing implement ation frameworks that can be used to implement modern applications. Contai ners provide those benefits without introducing the general cost of the vi rtualization.\n\n It exists lots of open-source configuration managers su ch as kubernetes\\[2\\]\, nomad\\[3\\]\, apache Mesos\\[4\\]\, or docker s warm\\[5\\] to define and keep a standard modular applications in a consis tent state and to dispatch its services on a cluster of remote nodes. Howe ver\, there is still a lack of adopted open-source solution for easily des igning and distributing a dataflow (*i.e. a complex directed-acyclic-graph of tasks for processing data*) on a distributed infrastructure to improve the performance of such applications. It exists some initial solution suc h as TaskCluster\\[6\\] from Mozilla dedicated to Continuous integration b ut it still remains the needs to design these data-flows and to efficientl y schedule the execution of its dependent tasks.\n\n Through this present ation/demo\, we are giving our experience feedbacks of building both a low cost and efficient transcoding platform over an ad hoc computing cloud bu ilt around a cluster of Raspberry Pis and containers solutions that use do cker\, hypriot / cluster labs \\[7\\] and go-docker \\[8\\]. We want in pa rticular discuss the need to provide better abstractions for easily managi ng hardware capabilities reservation such as GPU from the workflow.\n\n* 1 4:00 - Traefik\, un reverse-proxy/load-balancer dynamique écrit en GO\n **Emile Vauge** (Containous)\n\n* 14:45 - Piloter le réseau de vos conten eurs dans OpenStack avec Kuryr\n **Christophe Sauthier** (Objectif Libre) \n\n* 15:30 - Skydive: debugging and monitoring your networks from contain ers to the physical packets\n **Sylvain Afchain** (Red Hat)\n\n* 16:15 - PAUSE\n* 17:00 - How the OpenShift PaaS is implemented on top of Docker/Ku bernetes\n **Chmouel Boudjnah** (Red Hat)\n\n* 17:45 - Enjeux de l’orch estration des microservices\n **Hervé Leclerc** (Alter Way)\n\n* 19:00 - Conclusion et buffet-networking\n\n**[Inscription gratuite mais obligatoi re][3]**\n\nAutres journées à venir:\n\n [L'Agenda du Libre \\| Frama-C Day - Analyse et vérification de code ](../11226) [L'Agenda du Libre \\| Techniques de programmation web à l'état de l'art](../11227)\n\n[1]: htt p://www.open-source-innovation-spring.org/\n[2]: http://www.open-source-in novation-spring.org/open-source-pour-le-cloud-et-les-conteneurs/\n[3]: htt p://www.eventbrite.com/e/open-source-pour-les-conteneurs-et-les-microservi ces-tickets-23400662010\n LOCATION:23-25 avenue du Docteur Lannelongue\, Paris\, Île-de-France\, Fra nce ORGANIZER:mailto:muriel.shanseifan@systematic-paris-region.org SUMMARY:Open Source pour le Cloud et les Conteneurs URL;VALUE=URI:http://www.eventbrite.com/e/open-source-pour-les-conteneurs-e t-les-microservices-tickets-23400662010 X-ALT-DESC;FMTTYPE=text/html:
Quatrième événement du Printemps de l'innovation Open Source\, org anisé par le GTLL de Systematic et l'Irill\, présidé par Roberto Di Cos mo
\nOpen Source pour le Cloud et le s Conteneurs
\nProgramme dirigé pa r Frédéric Lepied (Red Hat) et Gilles Muller (Inria / LIP6 )
\nVivien Quéma (IMAG)\, Gaël Thomas (Telecom Sud-Paris)\, Pierre Sens (LIP6) \, Gauthier Voron (LIP6)
\nCloud providers are now proposing large virtua l machines with several tens of cores. Since such virtual machines often relies on complex hardware architectures\, new challenges arise for hyper visors to achieve good performances. This talk will present some of these problems and why they happen.
\nJulien Sopena (LIP6)
\nOlivier Barais et David Bromberg (IRISA).
\nThe increasing popularity of vid eos over Internet\, combined with the wide heterogeneity of various kinds of end users’ devices\, imposes strong requirements on the underlying in frastructure and computing resources to meet the users expectations. In pa rticular\, designing an adequate transcoding workflow in the cloud to stre am videos at large scale is: (i) costly\, and (ii) complex. Lightweight co ntainers technologies such as docker are more and more adopted for deployi ng and managing micro-services architectures [1]. Indeed\, they hide a par t of the complexity resulting from the language heterogeneity and the larg e choice of existing implementation frameworks that can be used to impleme nt modern applications. Containers provide those benefits without introduc ing the general cost of the virtualization.
\nIt exists lots of open -source configuration managers such as kubernetes[2]\, nomad[3]\, apache M esos[4]\, or docker swarm[5] to define and keep a standard modular applica tions in a consistent state and to dispatch its services on a cluster of r emote nodes. However\, there is still a lack of adopted open-source soluti on for easily designing and distributing a dataflow (i.e. a complex dir ected-acyclic-graph of tasks for processing data) on a distributed inf rastructure to improve the performance of such applications. It exists som e initial solution such as TaskCluster[6] from Mozilla dedicated to Contin uous integration but it still remains the needs to design these data-flows and to efficiently schedule the execution of its dependent tasks.
\nThrough this presentation/demo\, we are giving our experience feedbacks o f building both a low cost and efficient transcoding platform over an ad h oc computing cloud built around a cluster of Raspberry Pis and containers solutions that use docker\, hypriot / cluster labs [7] and go-docker [8]. We want in particular discuss the need to provide better abstractions for easily managing hardware capabilities reservation such as GPU from the wor kflow.
\nEmile Vauge (C ontainous)
\nChristophe Sauthier (Objectif Libre)
\nSylvain Afchain (Red Hat)
\nChmouel Boudjnah (Red Hat)
\nHervé Leclerc (Alter Way)
\nInscription gratuite mais obligatoire
\nAutres journées à venir :
\n