Workshop on Dependability and Interoperability in Heterogeneous Clouds (DIHC13)

in conjunction with

Euro-Par 2013 August 27th, Aachen, Germany

  • Contest competition

      The contest competition shall provide a forum to present and discuss the use of Contrail software: new applications using the Contrail software stack or some of its components, demonstration testbed of an existing application ported on Contrail, new tools that interoperate with and integrate Contrail software.
      Submissions should be in the form of a package including the code accompanied by two-pages proposal providing informal description of the demonstration (consisting of an application or tool), which parts of the Contrail software have been used, the experience in installing and using the Contrail software and advantages (if any) of using the Contrail software, what attendees will be able to see or do during the demo (e.g. screenshots). The same document must provide information about any equipment or facilities required for your system demonstration.
      All submissions will be evaluated mainly based on the quality of work, novelty of the application or tool, the potential to stimulate interesting discussions, exchange of ideas and promote collaborations.

      The winner of the competition will be invited to DIHC 2013 workshop to present the application/tool and work accomplished; the expenses will be partially covered.
      The winner will also be invited to submit a paper to be published in the post-proceedings of the workshop. More information will come after the workshop.

      The DIHC 2013 workshop aims at bringing together researchers from academia and industry interested in the design, implementation, and evaluation of services and mechanisms for dependable cloud computing in a multi-­cloud environment.

  • Contest competition important dates

      Registration and Code Submission: August 10, 2013
      Notification of the Winner: August 15, 2013
      Workshop date: August 27, 2013

  • Contest submission and support

      All the submissions should be addressed to contrail-dihc13-contest@inria.fr.
      Support will be provided to all participants of the Contrail contest; the support team can be contacted by email.

  • Contrail software

    Contrail brings a number of important features to Cloud administrators and users. With the Contrail software a Cloud administrator now can manage a Federation of Clouds providing support for basic SLA functionalities at the provider level on top of the deployment and management of the entire life-cycle of applications on a data center. A user accesses the Federation Web, which is completely integrated with authentication and authorization functionalities, meaning that user login obtains proper OAuth tokens that can be used for other calls. Applications are described via proper OVFs, where virtual networks can also be specified. A Contrail user can also access PaaS services via the ConPaaS component, designed to host both high-performance scientific applications and online Web applications. Contrail is available either as integrated software or as a set of individual components: ConPaaS (PaaS services), Security (authentication and authorization), SLA life-cycle management, VEP (data center and application management), VIN (virtual networks), GAFS (reliable distributed storage).

    The official deployment documentation for the Contrail can be found on the following link: http://contrail.projects.ow2.org/xwiki/bin/view/Documentation/WebHome. The page containing the list of binary packages with description of the most relevant API can be found here.
    The page contains links to the Installation Guide, which describes the default deployment process of a federation and provider software. The Administration Guide provides more detailed information about the alternative configurations of the components. Both guides are based on the deployment from the packaging system for either Ubutnu or Debian based distributions. The latest stable release of the software is found in the “release” repository, while daily builds can be found in the “testing” repository. More information on the packages, including tar.gz versions of all components are located at Contrail download page at http://contrail.projects.ow2.org/xwiki/bin/view/Main/Download.
    The code is hosted in the SVN on the OW2. Quick look of the code is available by using the web access at the following link: http://websvn.ow2.org/listing.php?repname=contrail.
    For people interested in the security aspects of the platform, we suggest to use the code from the branches. For example Federation-API and Federation-Web. These are part of the upcoming 1.3 release, currently in the beta-3 stage. To get a quick overview of the security aspects of the federation, have a look at the installation instructions at the wiki, as well as OAuth Authorization server description. Federation certificates usage is described here. If interested in work on security aspects, please, contact us at contrail-dihc13-contest@inria.fr.

  • Contrail components

    Contrail integrates different components. Herein we provide a brief descriptions and links to the components that can be used for this Contrail contest competition.

    ConPaaS (Contrail PaaS)

    ConPaaS (http://www.conpaas.eu/) software is at release ConPaaS-1.2.0. This release extends the previous versions and brings a number of important new features:

    • Multi-cloud deployments: ConPaaS can now be deployed over several clouds simultaneously. The user can choose in which cloud each virtual machine gets deployed. This may be useful for example to deploy ConPaaS over multiple availability zones in Amazon EC2, or to integrate resources from a private cloud together with a public one.
    • Virtual private network support: all communications between virtual machines belonging to the same application may now go through a multipoint virtual private network. This allows one to hide the network heterogeneity between machines located in different clouds, and to easily traverse firewalls. This feature exploits the IPOP technology.
    • Automated deployment of entire applications: instead of deploying multiple related ConPaaS services one by one, ConPaaS users can now deploy entire applications in one click. All that is necessary is to write a simple "application manifest" which specifies the list of required services and their configuration. ConPaaS provides a few ready-made manifests for common applications.
    • Performance monitoring: all ConPaaS services now have the ability to monitor their own performance using standard and service-specific metrics. ConPaaS is a runtime environment for hosting applications in the cloud. It aims at offering the full power of the cloud to application developers while shielding them from the associated complexity of the cloud. ConPaaS is designed to host both high-performance scientific applications and online Web applications. It automates the entire life-cycle of an application, including collaborative development, deployment, performance monitoring, and automatic scaling. Finally, it runs on a variety of public and private clouds, and is easily extensible. This allows developers to focus their attention on application-specific concerns rather than on cloud-specific details.

    Virtual Execution Platform (VEP)

    Virtual Execution Platform (https://project.inria.fr/vep) is a cloud middleware software that interfaces multiple Infrastructure as a Service (IaaS) clouds. VEP offers two types of services:

    1. management of data centres for a Cloud administrator allowing greater control on the physical resources;
    2. management of end-user distributed applications made up of several inter-networked virtual machines on a Cloud with an interface facilitating the deployment and application life-cycle management.

    The VEP software is at release v2.0 (download here), a new version replacing the old one (v1.1). The supported features are:
    • OpenNebula: VEP allows to manage a OpenNebula Cloud System offering a completely transparent interface;
    • Open Standard support: VEP respects the DTMF OVF and CIMI standards;
    • Data centre representation: VEP stores the provider datacenter layout to enforce some SLA requirements;
    • Physical resource allocation and resource provisioning for end user’s applications;
    • Easy interfaces: easy browser based cloud administrator and end-user application management interfaces;
    • VM Scheduling: the new VEP scheduler is independent from the IaaS Scheduler to allow placement constraints on the resources;
    • Advance Reservation: VEP allows to reserves a number of VMs from a start date to an end date to guarantee the deployment (in addition to run an application in best effort mode).

    VEP aims at offering the full management of IaaS cloud while shielding the administrators the complexity of managing heterogeneous resources. VEP is designed to provide interoperability by offering a uniform way of representing and managing the resources of a cloud provider. VEP could also easily enable the participation of a Cloud provider to a federation seamlessly and it does proper VM contextualization and application lifecycle management. Additionally it publishes application events and metrics for application’s monitoring and SLA enforcement.
    VEP enables interoperability through its RESTful interface based on the DMTF’s Cloud Infrastructure Management Interface (CIMI) standard. The CIMI model defines a framework for the application life cycle management on a cloud provider infrastructure, with applications generated from an Open Virtualization Format (OVF) document. VEP extends the CIMI API to support the deployment of applications under SLA terms.
    VEP RESTful interface allows cloud administrators to manage and control numerous aspect of the service offering. For VM scheduling to work properly, the datacenter topology information is desired in VEP. The software has a simple and intuitive interface to allow administrators input topology information about their datacenter. This interface also permits the administrator to selectively permit hosts to be managed via the VEP software. VEP never schedules a VM on a host not permitted for use by the administrator.
    An Installation Guide and a User Guide is availabe at the VEP project site https://project.inria.fr/vep/documentation/.

    Global Autonomous File System (GAFS) -- XtreemFS

    GAFS (http://www.xtreemfs.org/) is a cloud file systems built on XtreemFS which is at release v1.4. This release is the result of almost one thousand changes ("commits") to the code repository, and extensive testing throughout the year. We worked both on major improvements to the existing code and new features:

    • Improved stability: Clients and servers are rock solid now. In particular, we fixed client crashes due to network timeouts and issues with the Read/Write file replication.
    • Asynchronous writes: Once enabled (mount option "--enable-async-writes"), write() requests will be executed in the background. This improves the write throughput without weakening semantics. We recommend to enable async writes.
    • Windows Client (beta): Complete rewrite based on the stable C++ libxtreemfs and using the Dokan alternative Callback File System by EldoS corporation. Try it by mounting our public demo server!
    • Hadoop support: Use XtreemFS as replacement for HDFS in your Hadoop setup. This version of XtreemFS comes with a rewritten Hadoop client based libxtreemfs for Java which also provides data locality information to Hadoop.
    • libxtreemfs for Java: Access XtreemFS directly from your Java application. See the user guide for more information.
    • Vivaldi integration: The Vivaldi replica placement and selection policies enable clients to select close-by replicas based on actual network latencies. These latencies are estimated using virtual network coordinates which are also visualized in the DIR web-interface. Check out the demonstration on the web-interface of our public demo server.
    • Extended OSD Selection: Now you can assign custom attributes to OSDs and limit the placement of files on OSDs based on those attributes.

    This version also includes an updated version of the DIR/MRC replication and adds fail-over support for DIR replicas. As DIR/MRC replication is still in a very early stage this feature is intended as technology preview for more experimental users.

  • Contrail project

      Contrail (http://contrail-project.eu is a European project addressing interoperability and dependability to allow a federation of heterogeneous clouds to deploy distributed applications under QoS and QoP constraints.
      Contrail is an open source integrated approach to virtualization, offering Infrastructure as a Service services (IaaS), services for federating IaaS clouds and Contrail Platform as a Service services (ConPaaS) on top of federated clouds. The aim of the project is designing, implementing, evaluating and promoting an open source computational cloud wherein users can limitlessly share resources.

  •