Bench4Q is a QoS(Quality of Service)-oriented load testing tool. The goal of this topic is to improve the user experience of Bench4Q as a service. This has to be done by extending the test script design module,existing load injection module and service portal. A prototype and a tutorial will be provided.

The goal is to define a Rest API in PHP for Bitcoin and to integrate it on the emerginov platform. The emerginov platform is a PHP PaaS (Platform as a Service). The idea is to offer an API (and an associated SDK) to help developers of micro services to integrate a virtual money API based on Bitcoin.

On the emerginov PaaS (platform as a service) we used to manage lot of mobile micro services dealing with different topics (education, health, agriculture,..)
So far all the projects are listed see https://www.emerginov.org/projects_list.php.
Each project has tags declared by the project leader.
It is really messy..The goal is to develop an intuitive web portal able to visualize a list of projects. Based on criteria to be defined, the projects should be easily and graphically found. The criteria could be based on the number of lines, the activity, the tags, the number of commiters, a public rating system... 

They are lots of popular calendar application. And many mobile services need a back end calendar subscription/notification system. The work will consist in specifying, developing and testing an API (with an associated SDK) in PHP to allow the developers of emerginov microservice to use one of the popular calendar associated with a subscription/notification system (mail, SMS, IVR). This API will be integrated into the API shop of the open source emerginov solution hosted in OW2.

Description
MapReduce is a patented software framework introduced by Google to support distributed computing on large data sets on clusters of computers [1]. The framework is inspired by map and reduce functions commonly used in functional programming, although their purpose in the MapReduce framework is not the same as their original forms. MapReduce is a framework for processing huge datasets on certain kinds of distributable problems using a large number of computers (nodes), collectively referred to as a cluster. Computational processing can occur on data stored either in a filesystem (unstructured) or within a database (structured). The objective of this project is therefore to combines the Apache Hadoop implementation of MapReduce [2] and the FraSCAti platform [3] to support the development, deployment and execution of MapReduce applications.

References
[1] Google. The MapReduce Framework. http://en.wikipedia.org/wiki/MapReduce
[2] Apache Software Fundation. The Hadoop Framework. http://hadoop.apache.org
[3] OW2 FraSCAti project. http://frascati.ow2.org

Description
JVisualVM [1] is a visual tool integrating several commandline JDK tools and lightweight profiling capabilities. Designed for both production and development time use, it further enhances the capability of monitoring and performance analysis for the Java SE platform.
Application Developer: Monitor, profile, take thread dumps, browse heap dumps
System Administrator: Monitor and control Java applications across the entire network
Java Application User: Create bug reports containing all the necessary information
JVisualVM is based on an extensible architecture, which can be tuned in order to display application-specific or technology-specific information [2]. For example, an OSGi plugin has already been developed as part of the Chameleon project [3]. Therefore, the objective of this project is to develop a comprehensive plugin for the SCA component model and the FraSCAti platform. This plugin should provide the capability of introspecting an SCA application and to reconfigure its properties at run-time, similarly to the features currently available in the FraSCAti Explorer tool [4].

References
J. Sedlacek and T. Hurka. JVisualVM. http://visualvm.java.net/index.html
J. Sedlacek and T. Hurka. Getting Started Extending VisualVM. http://visualvm.java.net/api-quickstart.html
OW2 Chameleon project. JVisualVM plugin for OSGi. http://wiki.chameleon.ow2.org/xwiki/bin/view/Main/AdminTools
OW2 FraSCAti project. FraSCAti Explorer. http://frascati.ow2.org/doc/1.1.1/ch06.html

With the development of Web Service, there are a lot of online services, employing and developing proprietary technology to serve their users. Most of the services are single and independent. But there is a tendency to change the online services from single to composite. From the perspective of users, sometimes they don’t know how to choose from the numerous services. In order to help users make the correct choice form service provider quickly and conveniently, the system provides the function of service recommendation and returns to users a relatively optimized choices taking into account users’ non-functional requirement, such as personal privacy, searching habits and so on.

The subject will develop an Forge system which integrates the social tools with software development forges such as TrustieForge or other forges similar to TrustieForge. Such socialization may include:
 - blog or microblog
 - firends, groups, followers, "like", etc
 - social network analysis tools based on big repository data

Hive is a data warehouse system for Hadoop that facilitates easy data summarization, ad-hoc queries, and the analysis of large datasets stored in Hadoop compatible file systems.  However, compared with mature relational databases, Hive lacks several corresponding tools which is important to users, including performance and resource monitor, query planner, management platform and so on. HiveAdmin, the web management platform for Hive, aims at providing the facilities mentioned above.

A PaaS platform is usually expected to support multiple kinds of application servers so as to run various types of applications. The application servers can function well only if they are properly configured and well managed. Considering the large quantity of potential app servers in a PaaS platform, it is a heavy and inefficient work to manually configure app servers. The goal of this project is to design a uniform, convenient way to help the platform administrator manage all the app servers in Service4All. The contesters are required to fulfill the following tasks: (1) Extend Service4All to support Jboss or/and Weblogic. (2) Encapsulate the configuration and management functions of the extended app servers as Web services. (3) Implement the frequently-used configuration workflow with service composition so as to support advanced configuration work. (4) Design and implement corresponding web interfaces, and integrate with Service4All.

Service4All, as a cloud computing platform for service oriented software developers, has provided many services to facilitate the development, deployment, running and invocation of atomic web services, BPMN-based composite web services and java web applications. Although Service4All can replicate applications across servers, the SOAP layer of message transportation in Service4All may limit the performance. To improve the scalability of Service4All and achieve better load balance, Nginx, a widely-used reverse proxy, can be used by Service4All, which can dispatch high concurrent HTTP requests among replicas, so as to enable Service4All to deploy and run more atomic web services and java web applications. However, atomic web services and java web applications are dynamically deployed into Service4All and the current version of Nginx does not support the dynamic configuration of reverse proxy, therefore it is necessary to extend Nginx and the deployment/undeployment and replication components of Service4All to support dynamical reverse proxy for new web services and java applications. The objective of this project is to add the dynamic reverse proxy support to Service4All.

Service Oriented AppEngine is one pillar of Service4All and provides a hosting environment for atomic web services, BPMN-based composite web services, java web applications and other applications. Since a lot of different services and applications can be deployed on limited physical resources, the performance will be a critical factor. To improve the performance of AppEngine, we will employ the widely-used cache tools such as memcache to AppEngine, which can lower the latency of read operations. However, as hosted applications are massive and diverse, and the memory is volatile, we will build a shared distributed cache based on memcache for all applications if they need to cache their data and use replication to improve the reliability of data in the cache. The objective of this project is to add the distributed cache support to Service4All.

CRM is an important enterprise software, which is faced with scalability challenge when there are a large amount of users. Service4All is a PaaS platform, which can provide applications with highly scalable and available runtime environment. The objective of this project is to investigate the possibility of  combining the two opensourced work and to deliver a prototype of scalable CRM.

Efficient urban planning in today’s smart cities not only depends on the accessible physical infrastructure but also on the amount and quality of available information and communication. The CHOReOS middleware targets to support sensing of data from mobile devices on ultra large scales, towards providing access to information contributed by the participants. 

The project consists of using the CHOReOS middleware to design and develop innovative applications that the inhabitants of smart cities can use to improve their living. These applications will be based on data originating from sensors on Android devices (location, noise level, etc.) but furthermore, the use of data from non-sensing services that may exist in a smart city (buses schedule, gas prices, etc.) is also encouraged. Some examples include:  students looking for a quiet place to study before their exams, and a noisy, happening place for partying afterwards or (ii) people searching for the least crowded bus station nearby to catch their bus.

A secondary objective of this project will be to release the collected sensing data from the app as a publicly available dataset, to be used by researchers worldwide working on large scale urban mobile systems.

With the development of open source, there are more and more open source projects on the Internet. However, it is time-consuming for developers to master an unfamiliar open source project or API due to large amount of source code and documents.
Nowadays, developers usually use search engines like Google to find some information when they have questions about open source projects. But search engines cannot comprehend the semantic information in queries. And retrieval just by keyword matching often mixes results with many noisy data.
The goal of this project is to implement a tool that can extract and organize useful information from existing source code and documents, and then combine NLP technologies to support natural language queries.

In software development, reusing existing Application Programming Interfaces(APIs) significantly improves programmer productivity and software quality.
Code examples are important and key resource for understanding the usage of APIs. However, one study found that the greatest obstacle to learning an API in practice is "insufficient or inadequate examples". It tends to be time-consuming and require significant human effort for developers searching API information, such as examples, on the Web to understand the right usage of APIs.
Many existing work mines API usage examples from client code repositories. But examples from client codes are always long and complex, also the client code cannot be obtained when projects newly publish.
The goal of this topic/project is to mine efficient and easy-to-understand usage examples for APIs in newly published projects.

•Academic research on open source software is of great value.
•These research work often requires a significant amount of open source software resources of various kinds(code, documentations, issue tracker, mailing list…), which are widely spread on the Internet and are of vast quantity.
•These resources are mostly manually searched and downloaded from the Internet, which is time consuming. On the other hand, it is hard to manage, to maintain and to update the downloaded resources.
•The goal of this project is to implement a tool that can automatically find, download, manage open source software resources, which helps the researchers improve their efficiency in academic research.

ProActive Parallel Suite is an Open Source middleware featuring Programming, Scheduling and Resourcing tools for parallel, distributed and multi-core computing (http://proactive.activeeon.com/index.php).

Jobs describing the tasks to be ran are submitted to the Scheduler for distributed execution. Today these jobs can be written manually in XML or built through an GUI using Eclipse RCP technology (http://hudson.activeeon.com/view/RCP/job/Workflow_Studio/lastSuccessfulBuild/artifact/com.activeeon.workfloweditor.help/doc/built/QuickStart/single_html/QuickStartManual.html). We would like to build a new web application: the Web Workflow Studio for ProActive.

The top features would be:
- creating and editing task workflows with looping and branching
- editing tasks and jobs properties, including source code editing within the browser (for part of the jobs that are scripts)
- submitting jobs through the Scheduler's REST API (https://portal.cloud.sophia.inria.fr/SchedulingRest/)
- importing existing jobs & exporting jobs (http://hudson.activeeon.com/view/Scheduling/job/Scheduling_3.3.x/lastSuccessfulBuild/artifact/doc/built/Scheduling/single_html/ProActiveSchedulerManual.html#XSDJobSchema)

The project can be split in two phases:
- analyse potential solutions to build such web application, our existing web applications (https://portal.cloud.sophia.inria.fr/scheduler/ &  https://portal.cloud.sophia.inria.fr/rm/) are using GWT and SmartGWT but we are open to pure JS libraries.
- build a prototype with the chosen solution

You should be familiar with web technologies and will be helped by our software engineers from Sophia Antipolis, France (http://www.activeeon.com/company).

The CHOReOS middleware enables People, Services and Things to be connected and interact in numerous ways, by seamlessly integrating the heterogeneous interaction paradigms involved, such as tuple space, client-service and publish-subscribe. Software entities employing these paradigms can be deployed and run on top of the customized CHOReOS ESB.
 
The proposed topic consists in using the CHOReOS middleware to design and develop software for sensors deployed in a city, services processing sensor data, and event-based mobile user applications for smart cities.

More specifically, students will be able to develop sensor software that collect data for weather forecast, traffic estimation and air pollution, and store them in a shared data space. These data are processed by specialized services, and the resulting information is provided to subscribed end users on their mobile application. At the same time, end users are able to share their own feeling about city conditions.

Btrplace (http://btrp.inria.fr) is the flexible VM placement algorithm inside Entropy. It allows to compute the VMs placement, the servers' state, the resource allocation with regards to high-level constraints expressed by users. The btrplace sandbox (http://btrp.inria.fr:8080/sandbox/) is used for visualization and demonstration purpose of the btrplace inferring capabilities.

The current version of the sandbox relies on an outdated version of btrPlace that prevent it to use the whole set of supported constraints. It also has limited edition and integration capacities. We want to get rid of these limitations therefore, the objective of this project is to update the sandbox code, and BtrPlace if necessary. In addition, we think the GUI should be improved to provide a better visualization of a reconfiguration process.

Required technical skills: Java, Javascript, HTML/CSS

The goal is to define a methodology and a prototype to manage what-if simulation in an abstract way, independently from a single business domain. That means to be able to define (thanks to metadata) what is the process you want to simulate, what are the control levers the users have, what are the rules to generate the new results. Finally, a prototype of an application where the end-user formulates new hypothesis on his business and evaluate their impacts is required.

Design and prototype a component able to manage multiple and eterogeneous data sets, enabling a cross-query without using a rdbms as back-end.
Data sets can come from different data sources (rdbms, file, Web  Services, etc)

Design and prototype an analytical front-end able to dinamically explore big-data thanks to a semantic layer that describes their meaning 

DocDokuPLM is Product Lifecycle Management solution constituted of a Java EE back-end and a front-end that makes intensive use of HTML5 technologies.
With such amount of JavaScript (BackboneJS framework) files, it seems to be crucial to define a solid test strategy.
That's the purpose of this project.
The volunteer will define and implement unit and functional testing suites using the latest JavaScript frameworks like Jasmine, NodeJS, PhantomJS...

The goal of the service cloud is to leverage the Petals Service Bus to the cloud and to create what will be called the Service Cloud: Bringing SOA and integration services as a Cloud Service (iPaaS).
The work will focus on how to extend the Petals kernel and on developing new cloud-focused controllers to manage the distributed runtime on multiple IaaS.
Implementation: Java, Web services (REST), OpenStack.