JUNIPER Project2024-03-29T14:49:37ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeidahttp://storage.ning.com/topology/rest/1.0/file/get/2996391079?profile=RESIZE_48X48&width=48&height=48&crop=1%3A1http://www.juniper-project.org/forum/topic/listForContributor?user=0g9x6jhcj7uk5&feed=yes&xn_auth=noJUNIPER results availabletag:www.juniper-project.org,2016-02-04:6620280:Topic:44112016-02-04T17:50:21.225ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>The public technology results from the JUNIPER project have been made available for download. The complete software stack developed within the project is available for evaluation. Also, the public technical reports can downloaded. Click on the <a href="http://www.juniper-project.org/page/publications" target="_blank">Results</a> tab at the top of the page for access. </p>
<p>The public technology results from the JUNIPER project have been made available for download. The complete software stack developed within the project is available for evaluation. Also, the public technical reports can downloaded. Click on the <a href="http://www.juniper-project.org/page/publications" target="_blank">Results</a> tab at the top of the page for access. </p> Tutorial on JUNIPER programming model published on Modelio communitytag:www.juniper-project.org,2016-01-20:6620280:Topic:42032016-01-20T16:15:25.322ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>This tutorial presents the programming model put forward by JUNIPER and realized in the form of the JUNIPER MDE Environment. To illustrate the approach we will build a simple data streaming pipeline based on the well-known map-reduce pattern.</p>
<p>For more information, <a href="https://www.modelio.org/forum/23-experiments/3867-models-for-big-data-stream-processing.html#3867" target="_blank">click here</a>.</p>
<p>This tutorial presents the programming model put forward by JUNIPER and realized in the form of the JUNIPER MDE Environment. To illustrate the approach we will build a simple data streaming pipeline based on the well-known map-reduce pattern.</p>
<p>For more information, <a href="https://www.modelio.org/forum/23-experiments/3867-models-for-big-data-stream-processing.html#3867" target="_blank">click here</a>.</p> JUNIPER Meeting in Softeam Cadextan in Paristag:www.juniper-project.org,2015-11-05:6620280:Topic:39032015-11-05T13:32:40.672ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p><span style="text-decoration: line-through;"><em><span class="font-size-6"><a href="http://storage.ning.com/topology/rest/1.0/file/get/2869436840?profile=original" target="_self"><img class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869436840?profile=RESIZE_1024x1024" width="750"></img></a></span></em></span> <a href="http://storage.ning.com/topology/rest/1.0/file/get/2869443683?profile=original" target="_self"><img class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869443683?profile=RESIZE_1024x1024" width="750"></img></a> <a href="http://storage.ning.com/topology/rest/1.0/file/get/2869444474?profile=original" target="_self"><img class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869444474?profile=RESIZE_1024x1024" width="750"></img></a></p>
<p><span style="text-decoration: line-through;"><em><span class="font-size-6"><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869436840?profile=original"><img width="750" class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869436840?profile=RESIZE_1024x1024" width="750"/></a></span></em></span><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869443683?profile=original"><img width="750" class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869443683?profile=RESIZE_1024x1024" width="750"/></a><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869444474?profile=original"><img width="750" class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869444474?profile=RESIZE_1024x1024" width="750"/></a><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869444872?profile=original"><img width="750" class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869444872?profile=RESIZE_1024x1024" width="750"/></a></p> New Draft Standard for Realtime Javatag:www.juniper-project.org,2015-02-05:6620280:Topic:29052015-02-05T15:40:54.231ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<div class="content"><p>JSR 282, the Realtime Specification for Java 2.0, has just released a<br></br> draft of the specification for community review. The draft can be found<br></br> at <a class="moz-txt-link-rfc2396E" href="https://www.jcp.org/en/jsr/detail?id=282" rel="nofollow">https://www.jcp.org/en/jsr/detail?id=282</a>. In case you have<br></br> interest in realtime Java, please review the specification. I would be<br></br> happy to answer any questions you may have.</p>
<p>Dr. James J. Hunt * CEO &…</p>
</div>
<div class="content"><p>JSR 282, the Realtime Specification for Java 2.0, has just released a<br/> draft of the specification for community review. The draft can be found<br/> at <a rel="nofollow" class="moz-txt-link-rfc2396E" href="https://www.jcp.org/en/jsr/detail?id=282">https://www.jcp.org/en/jsr/detail?id=282</a>. In case you have<br/> interest in realtime Java, please review the specification. I would be<br/> happy to answer any questions you may have.</p>
<p>Dr. James J. Hunt * CEO & Geschäftsführer *</p>
<p>aicas GmbH<br/> Haid-und-Neu-Straße 18 * D-76131 Karlsruhe * Deutschland (Germany)<br/> <a rel="nofollow" class="moz-txt-link-freetext" href="http://www.aicas.com">http://www.aicas.com</a> * Tel: +49 721 663968 0 * FAX: +49 721 663968 99</p>
</div> BigDataComputing workshop organized by JUNIPER (in collocation with EuroMPI/ASIA international conference)tag:www.juniper-project.org,2015-01-05:6620280:Topic:24222015-01-05T09:52:08.035ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p><a href="http://storage.ning.com/topology/rest/1.0/file/get/2869436357?profile=original" target="_self"><img class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869436357?profile=original" width="259"></img></a> The workshop took place on 10 September 2014 during the EuroMPI/ASIA conference as a 1-day event aiming to bring together researchers from the domains related to data-centric computing for discussion on challenges and possible solutions for <strong>data-centric application development in high-performance computing environments</strong>. The workshop has resulted in a…</p>
<p><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869436357?profile=original"><img class="align-full" src="http://storage.ning.com/topology/rest/1.0/file/get/2869436357?profile=original" width="259"/></a>The workshop took place on 10 September 2014 during the EuroMPI/ASIA conference as a 1-day event aiming to bring together researchers from the domains related to data-centric computing for discussion on challenges and possible solutions for <strong>data-centric application development in high-performance computing environments</strong>. The workshop has resulted in a well-attended and highly successful dissemination event, which sparked a decent outcry in scientific and industrial communities attending the renomated EuroMPI conference.</p>
<p>In particular, the JUNIPER project's outcomes were presented with the papers of Ian Gray et al. "Architecture-Awareness for Real-Time Big Data Systems" and Alexey Cheptsov "HPC in Big Data Age: An Evaluation Report for Java-Based Data-Intensive Applications Implemented with Hadoop and OpenMPI".</p>
<p>The workshop's proceedings can be found at the <a href="http://dl.acm.org/citation.cfm?id=2642769&picked=prox" target="_blank">Link</a>.</p> Workshop on Real-Time Scheduling in the Linux Kernel (RTS-LIKE)tag:www.juniper-project.org,2014-06-25:6620280:Topic:15042014-06-25T09:26:01.292ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>As the new scheduling class SCHED_DEADLINE entered the Linux kernel, new opportunities are available to time-sensitive applications on Linux.<br></br>The RTS-LIKE workshop aims at establishing a fruitful and open forum between Linux kernel developers and real-time researchers.<br></br>The workshop program is composed by both talks on real-time scheduling issues and presentations by real-time applications programmers, reporting their experiences and use-cases.</p>
<p>Website: …</p>
<p>As the new scheduling class SCHED_DEADLINE entered the Linux kernel, new opportunities are available to time-sensitive applications on Linux.<br/>The RTS-LIKE workshop aims at establishing a fruitful and open forum between Linux kernel developers and real-time researchers.<br/>The workshop program is composed by both talks on real-time scheduling issues and presentations by real-time applications programmers, reporting their experiences and use-cases.</p>
<p>Website: <a href="http://retis.sssup.it/rts-like/" target="_blank">http://retis.sssup.it/rts-like/</a></p> SCHED_DEADLINE in mainlinetag:www.juniper-project.org,2014-02-12:6620280:Topic:12822014-02-12T14:21:48.105ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>Most RTOSs provide <a href="http://en.wikipedia.org/wiki/Fixed-priority_pre-emptive_scheduling">Fixed Priority scheduling</a> as the only available scheduling policy. This is because FP is effective, and easy to be implemented efficiently. Also, FP is in the POSIX standard, so all UNIX-based systems implement it. Different flavours of FP are also available in Windows and Mac OS.</p>
<p>At the <a href="http://retis.sssup.it/">RETIS lab</a>, we have based much of our research on an alternative…</p>
<p>Most RTOSs provide <a href="http://en.wikipedia.org/wiki/Fixed-priority_pre-emptive_scheduling">Fixed Priority scheduling</a> as the only available scheduling policy. This is because FP is effective, and easy to be implemented efficiently. Also, FP is in the POSIX standard, so all UNIX-based systems implement it. Different flavours of FP are also available in Windows and Mac OS.</p>
<p>At the <a href="http://retis.sssup.it/">RETIS lab</a>, we have based much of our research on an alternative scheduler, <a href="http://en.wikipedia.org/wiki/Earliest_deadline_first_scheduling">Earliest Deadline First</a> (EDF), known to be a theoretically optimal scheduler. As part of these research efforts, the group at RETIS Lab of Scuola Superiore Sant'Anna, together with the people at <a href="http://www.evidence.eu.com/it">Evidence</a>, implemented some time ago the <a href="http://en.wikipedia.org/wiki/SCHED_DEADLINE">SCHED_DEADLINE scheduler</a>, while working on the <a href="http://www.actors-project.eu/">ACTORS EU project</a>. The scheduler has been continuously improved and maintained by our group over the years, and it is now being used in the <a href="http://www.juniper-project.org/" target="_blank">JUNIPER project.</a> <a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869442138?profile=original"><img width="300" class="align-right" src="http://storage.ning.com/topology/rest/1.0/file/get/2869442138?profile=RESIZE_320x320" width="300"/></a><a href="http://www.juniper-project.org/" target="_blank"></a></p>
<p>The SCHED_DEADLINE patch implements the EDF with the <a href="http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=739726&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D739726">Constant Bandwidth Server</a> (CBS) algorithm. According to this policy, each thread is assigned a budget Q and a period P. The thread is then guaranteed to receive Q units of execution time every P. Therefore, this policy combines real-time behavior (thanks to the EDF real-time policy) and temporal isolation (thanks to the budgeting policy). In particular, a thread cannot "disturb" the execution of the other threads in the system, because it cannot execute more than its assigned share of CPU Q/P. The scheduler is ready for multicore systems and it can support global as well as clustered and partitioned scheduling.</p>
<p>A few weeks ago, the scheduler went into the <a href="https://www.kernel.org/" target="_blank">mainline Linux kernel</a>. <a href="http://www.phoronix.com/scan.php?page=news_item&px=MTU3Njk">Here is one announcement</a>. Many thanks first of all to <a href="http://www.linkedin.com/profile/view?id=1383523&authType=NAME_SEARCH&authToken=iPDg&locale=en_US&srchid=315045201390255561768&srchindex=1&srchtotal=23&trk=vsrp_people_res_name&trkInfo=VSRPsearchId%3A315045201390255561768%2CVSRPtargetId%3A1383523%2CVSRPcmpt%3Aprimary">Peter Zjilstra</a> who has always supported this idea; and to <a href="http://www.osadl.org/Thomas-Gleixner.hannover-2008-thomas-gleixner.0.html">Thomas Gleixner</a> for his initial help and support all along the way. Also, many thanks to all the people that have contributed to its development: Michael Trimarchi, Claudio Scordino, Fabio Checconi, Luca Abeni, Dario Faggioli and Juri Lelli.</p>
<p>If you are interested, here are two recent papers on this scheduler.</p>
<ul>
<li>Juri Lelli, Giuseppe Lipari, Dario Faggioli, Tommaso Cucinotta, <i>An efficient and scalable implementation of global EDF in Linux</i> In proceedings of the 7th Workshop on Operating Systems Platforms for Embedded Real-Time Applications (OSPERT), Porto (Portugal), July 2011. (<a href="http://retis.sssup.it/%7Efaggioli/papers/OSPERT-2011-gedf.pdf">pdf</a>)</li>
</ul>
<ul>
<li>A. Parri, J. Lelli, M. Marinoni and G. Lipari. An implementation of the Bandwidth Inheritance Protocol in the Linux Kernel. In RTLWS'13. 2013. (<a href="http://retis.sssup.it/~lipari/papers/rtlws2013.pdf">pdf</a>)</li>
</ul> Juri Lelli invited at LKS 2013tag:www.juniper-project.org,2013-11-15:6620280:Topic:10932013-11-15T14:56:27.072ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p><strong>Juri Lelli</strong>, PhD student at Scuola Superiore Sant'Anna and maintainer of the <strong>SCHED_DEADLINE patchset</strong> for the Linux kernel, has been invited to the annual <strong>Linux Kernel Summit</strong>. The SCHED_DEADLINE patchset is part of a set of scheduling algorithms and kernel mechanisms available for Linux to enforce <strong>temporal constraints</strong> in multi-core systems. SCHED_DEADLINE is being developed and maintaned as part of Workpackage 3, System…</p>
<p><strong>Juri Lelli</strong>, PhD student at Scuola Superiore Sant'Anna and maintainer of the <strong>SCHED_DEADLINE patchset</strong> for the Linux kernel, has been invited to the annual <strong>Linux Kernel Summit</strong>. The SCHED_DEADLINE patchset is part of a set of scheduling algorithms and kernel mechanisms available for Linux to enforce <strong>temporal constraints</strong> in multi-core systems. SCHED_DEADLINE is being developed and maintaned as part of Workpackage 3, System Software, used inside the JUNIPER project to enforce predictability. More information about LKS can be found <a href="http://events.linuxfoundation.org/events/linux-kernel-summit" target="_blank">here</a>, while the ongoing discussion on the Linux Kernel Mailing List can be reached from <a href="https://lwn.net/Articles/573233/" target="_blank">here</a>.</p>
<p><a target="_self" href="http://storage.ning.com/topology/rest/1.0/file/get/2869436033?profile=original"><img class="align-center" src="http://storage.ning.com/topology/rest/1.0/file/get/2869436033?profile=original" width="480"/></a></p> New paper on Resource Reservation at the 15th RTLWStag:www.juniper-project.org,2013-11-15:6620280:Topic:11842013-11-15T14:53:05.667ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>At the <strong>15th Real-Time Linux Workshop</strong> held in Lugano-Manno on October 28-31, Andrea Parri and Juri Lelli of the ReTiS Lab of the Scuola Superiore Sant'Anna presented a paper describing the Multiprocessor Bandwidth Inheritance. M-BWI is a OS protocol used to improve isolation among resource reservations while dealing with mutex semaphores on shared resources. The work has been carried out in the context of the WorkPackage 3 of the JUNIPER project, concerning System Software.…</p>
<p>At the <strong>15th Real-Time Linux Workshop</strong> held in Lugano-Manno on October 28-31, Andrea Parri and Juri Lelli of the ReTiS Lab of the Scuola Superiore Sant'Anna presented a paper describing the Multiprocessor Bandwidth Inheritance. M-BWI is a OS protocol used to improve isolation among resource reservations while dealing with mutex semaphores on shared resources. The work has been carried out in the context of the WorkPackage 3 of the JUNIPER project, concerning System Software. More information about the workshop can be found <a href="https://www.osadl.org/Lugano-Manno-2013.rtlws15-lugano-manno-2013.0.html" target="_blank">here</a>, while the full paper can be downloaded from <a href="http://retis.sssup.it/~jlelli/papers/Design-Implementation-MBWI-Parri.pdf" target="_blank">here</a>.</p> New paper of A. Cheptsov and B. Koller "Message-Passing Interface for Java Applications: Practical Aspects of Leveraging High Performance Computing to Speed and Scal Up the Semantic Web" at IJAStag:www.juniper-project.org,2013-08-12:6620280:Topic:9732013-08-12T09:08:01.667ZMarcos Almeidahttp://www.juniper-project.org/profile/MarcosAlmeida
<p>The <strong>International Journal on Advances in Software (IJAS)</strong> has recently published a paper of <span style="text-decoration: underline;">Alexey Cheptsov and Bastian Koller (HLRS)</span> entitled "<em>Message-Passing Interface for Java Applications: Practical Aspects of Leveraging High Performance Computing to Speed and Scale Up the Semantic Web</em>". The paper discusses a strategy of using the Message-Passing Interface to parallelize data-centric Semantic Web applications. The…</p>
<p>The <strong>International Journal on Advances in Software (IJAS)</strong> has recently published a paper of <span style="text-decoration: underline;">Alexey Cheptsov and Bastian Koller (HLRS)</span> entitled "<em>Message-Passing Interface for Java Applications: Practical Aspects of Leveraging High Performance Computing to Speed and Scale Up the Semantic Web</em>". The paper discusses a strategy of using the Message-Passing Interface to parallelize data-centric Semantic Web applications. The study conducted by the authors has proved a high potential of MPI to be used as a core parallelization technology in JUNIPER as well, along with MapReduce/Hadoop. The full text can be found here: <a href="http://www.thinkmind.org/index.php?view=article&articleid=soft_v6_n12_2013_4" target="_blank">link</a>.</p>