USVI Public Records

Document text

10.4 Data Management Plan UVI’s Information Technology Services & Institutional Assessment (ISIA) and VI-EPSCoR have collaborated since 2003 to help fund critical infrastructure improvements. The partnership has successfully completed key projects including the installation of a 300-megabit Aviat Networks carrier- grade, wireless backhaul solution to connect St. Thomas and St. Croix campus networks and re- engineered UVI campus networks to accommodate high-speed Internet and Internet2 broadband connections. The recent RII Mare Nostrum project invests in UVI’s network infrastructure by upgrading core network switch and routing equipment from 1 Gbps to 10 Gbps, one of several recommendations in a network assessment performed by Sanforce LLC of Doral, Florida. Recent network test assessments prove the growth of VI-EPSCoR related research data transmission and Internet bandwidth needs continue to stretch the infrastructure capacity of the University and the Territory. Cyberinfrastructure’s mission remains to deliver to each research desktop 10 Gbps network capacity to store, calculate and present research data in a secure, resilient and redundant network environment. Cyberinfrastructure seeks to capitalize on past EPSCoR grant funding by • building upon core network infrastructure upgrades from 1 Gbps to 10 Gbps to research facilities; • improving the research data server and storage topology; • configuring the current MS Azure environment to replicate with a newly proposed on-premises data server and storage solution; and • improving University Internet connectivity data transmission throughput and speed. The following is a breakdown of the goals and objectives into five major tasks. 1. Multi-home the internet connection and increase bandwidth capacity to 10 Gbps on and between both campuses 2. Upgrade fiber optic infrastructure and pathway system for research buildings from multimode to singlemode fiber to support 10+ Gbps building uplink speeds 3. Upgrade research- related in-building wired and wireless services 4. Expand public cloud services in support of research data needs 5. Deploy on-premises research virtual server and storage environment with replication to MS Azure, and possibly other public cloud environments Data Management: VI- EPSCoR will implement a hybrid solution of on-premises virtual servers and storage that replicate to cloud environments to host its data. VI-EPSCoR research teams will have the option to locally store data in UVI’s on-premises data storage solution, which will replicate to MS Azure, allowing researchers local access during possible Internet outages. Both the on-premises and cloud solution will use a high- performance computing (HPC) operating system to store and manage research data files. MS Azure cloud platform provides VI-EPSCoR with integrated tools, development and operations, and management support. MS Azure also emphasizes hybrid databases that would combine research data on a University’s premise that can work in combination with the data in the cloud through a data sync tool. 1. Types of Data Produced - The majority of research data collected are physical oceanographic data (current speed/ direction, water temperature, dissolved oxygen, salinity), biological oceanography data (eggs and larvae of marine animals), acoustic telemetry data (animal movements), modelling data (model Page ! of ! 1 2 outputs based on physical and biological oceanographic data), soil samples (eroded soils on land and sedimentation in the marine environment, nutrients, DNA), and biological samples (fish (reproductive structures and stomach contents), algae, corals, plankton, DNA). Researchers will use 20 year time series benthic habitat data for analysis and modeling and will collect archival data on historical distribution of mangrove habitats. Other types of data collected, are EWFD curriculum materials and survey and analysis data. 2. Data and Metadata Standards - Most research data and its metadata will be comprised of two formats – contextual information about the data in a text-based document and SQL file formats. Other data formats will be vector or raster data sets stored as, e.g. GeoTiff, NetCDF, etc. for geospatial analysis. 3. Access and Sharing - Only authorized VI-EPSCoR users, tied to their roles and responsibilities, will have secure access through a web-based interface where it can be visualized and queried. The web application file format will be ASP.NET. Web accessible data will be discoverable via the use of persistent identifiers to reference webpages. Sharing of the data will be based on policy and procedures put forth by VI-EPSCOR PI and Co-PIs. 4. Re-Use, Distribution - Access to VI-EPSCoR research databases and associated web-based software tools generated under the project will be available for educational, research, and non-profit purposes. Such access will be provided using web-based applications, as appropriate. Materials generated from this project will be disseminated in accordance with the University, collaborating institutions, and NSF policies. This information will be described in the metadata and will be used as part of the outreach and knowledge translation portion of the project. Intended and foreseeable users of the data are EPSCoR jurisdictions and collaborating institutions. Those that use the data (as opposed to any resulting manuscripts) should cite it accordingly. 5. Archiving and Preservation - VI-EPSCoR research data will be preserved for the long-term in both the on-premises data store and in the Azure’s cloud environment. An abstract will be submitted with the datasets that describe their original context and potentially relevant project information. The abstract will make it easier to reference archived project data. UVI will be responsible for preparing data for long-term preservation and disaster recovery. Preservation of biological data samples will follow traditional sample protocols such as freezing of tissue, use of ethanol preservation, and safety and quality assurance protocols associated with the transport to off- island laboratories. Data Security: VI-EPSCoR research data will be secured with the latest privacy and compliance features within cloud-based technology and under current industry standards through the ability to manage and control identity and user access. VI-EPSCoR researchers and data participants will use their University login credentials for both local and cloud authentication and will have access in accordance with their roles and responsibilities. All VI-EPSCoR research data will be encrypted on its servers as data is inputted or uploaded and in its data storage to prevent cyber hacking. The Network Director and Network Engineer identified for this project will have protected administrative access to manage VI-EPSCoR research data’s on-premises and cloud-based server environment. All human subject research data will adhere to current IRB, HIPPA, and/or FERPA policies and restrictions. Page ! of ! 2 2