On October 7, 2008 we held the First Annual OSDT Workshop in La Jolla CA.

The intellectual origins of DataTurbine extend back to 1980 (see Matt Miller’s presentation for details). The motivation was provided by NASA applications that required real-time data management from a wide variety of sensors and instruments. Engineers from NASA and Creare Inc. continued to evolve the concepts and early software prototypes for the next 20 years. In 2001 these efforts coalesced into the DataTurbine Ring Buffered Network Bus (RBNB) invention, which was awarded US Patent #6,212,568.

During the period from 2001 until 2007, DataTurbine was a proprietary software product owned by Creare, Inc. It was developed as generic streaming data middleware for real-time data acquisition systems. It was successfully deployed in science and engineering applications funded by NSF, NASA, NOAA, and private industry. As the code base matured and the user and developer communities grew, the recognition of it’s utility spread. The evolution of DataTurbine coincided with advances in sensing and communications technologies and a desire by the science and engineering communities to deploy real-world large-scale sensor networks and environmental observing systems. In the fall of 2006 a small group of DataTurbine enthusiasts came together at UCSD to discuss the challenges faced by real-time systems developers and the options for making DataTurbine more accessible. It was noted that (1) there were many projects, extant and in planning, that could benefit from streaming data middleware, (2) DataTurbine satisfied many of the requirements for a robust streaming data middleware product, and (3) the proprietary status of the DataTurbine code limited it’s adoption by a broader community.

These observations lead to a plan: migrate DataTurbine from a proprietary software product into an open source software initiative. This plan was initiated with a successful proposal to the NSF Office of Cyberinfrastructure.

The OSDT Initiative was launched in October 2007 with a grant from the National Science Foundation Office of Cyberinfrastructure (award #OCI-0722067).This award funded the core activities needed to build an open-source software community around DataTurbine. There were three areas of funded activities: 1. Publish DataTurbine as an open source software product and provide developer support, including documentation, bug tracking, collaboration tools, and experimental facilities. 2. Enhance the code base, including porting DataTurbine to additional compute platforms, writing additional device drivers, and testing and tuning under real-world scenarios. 3. Build an active open source community through education, outreach, recruitment, and technical support.

At the one-year anniversary of the OSDT Initiative we held the first OSDT Initiative Workshop to share experiences and ideas, and to plan for future activities. The workshop was organized into four sessions. 1. Invited presentations on OSDT technology and applications. The primary objective was to hear from OSDT developers and users, in particular the types of deployments (e.g., science topics, types of sensors and instruments, and networking infrastructure), their experiences in using DataTurbine (e.g., usability, performance, robustness), and their ideas for new OSDT developments and activities. There were 12 presentations during this session on a broad range of applications, from environmental monitoring to precision engineering. 2. Presentations and discussions on DataTurbine deployment issues, including state of health monitoring, metadata management, time synchronization, networking, data replication and mirroring, and system configuration and management. 3. Presentations and discussions on DataTurbine software extensions. Among the topics discussed were the following: GoogleEarth KML plugins for OSDT, LabView interface to OSDT, OSDT support for GOES satellite imagery (www.goes.noaa.gov), and Google Protocol Buffers for OSDT. 4. Discussion of OSDT code management practices and developer support. This session reviewed the current open source support provided to the OSDT community, including code publication, quality control, bug tracking, technical consulting, and discussion forums. The OSDT Gforge system (http://gforge.org/) for code management was presented as well as the OSDT activities with the NSF NMI Build and Test Facility (http://nmi.cs.wisc.edu/) . In each session, items were identified for both immediate and longer-term action. These will be summarized and distributed as part of the next phase OSDT Initiative plan. For now the best source of meeting details are (1) the on-line presentations and abstracts, (2) the OSDT FAQ page, and (3) the OSDT developers’ mailing list (all links are provided above).

By all measures, the OSDT Initiative achieved the primary objectives and has reached critical mass in terms of community commitment. The four elements necessary for any successful open source software community are the following: 1. Reliable code: software that is useful and easy to use. 2. Applications and deployments: actual systems that employ the code and provide the application pull. 3. Developer support: documentation, bug tracking, technical consulting, and training. 4. An active and engaged user community: the human resources needed to turn physical resources into systems and solutions.

Catalyzed by the NSF award, the OSDT Initiative has been successful in pulling all of these elements together. The result is an active community of scientists and developers collaborating to produce useful middleware and successful deployments. Only one year since inception, the OSDT Initiative has demonstrated broad impact on a variety of projects and communities.

In summary, on 7 October 2008 we held the First Annual Open Source DataTurbine workshop. There were 25 participants from a number of science and engineering disciplines. The presentations and discussions of the workshop participants were lively and insightful. They provide critical guidance in planning the next phase of the OSDT Initiative. We concluded the workshop with renewed enthusiasm and optimism for the vision embodied in the original OSDT Initiative proposal. Real-time data applications, and sensor networks and observing systems in particular, are critical for many science and engineering domains. This continues to be an area of rapid growth. The Open Source DataTurbine Initiative provides a reliable code base and a supportive community for enabling these applications. We expect the second year of the Initiative to be even better than the first.

We would like to thank all of our members and our supporters.

Agenda, abstracts, and presentations: http://www.dataturbine.org/content/first-annual-open-source-dataturbine-...

Workshop participants: http://www.dataturbine.org/content/first-annual-open-source-dataturbine-...

Many of the workshop topics continue to be discussed on the developers mailing list: https://lists.sdsc.edu/mailman/listinfo/rbnb-dev

Many of the technical topics are summarized in the FAQ section of the OSDT website: (www.dataturbine.org/content/dataturbine-frequently-asked-questions).

Please refer to our paper titled "The Open Source DataTurbine Initiative: Streaming data middleware for environmental observing systems" (http://dataturbine.org/biblio/) for further details.