At SciServer, our core goals are to revolutionize the availability and accessibility of large-scale data-intensive science to the scientific community. Our goals encompass and build upon our joint history with SkyServer at Johns Hopkins University as we develop SciServer to provide the same unique capabilities across the scientific spectrum.
Building upon our decade-long effort on SDSS, SkyServer, and its ad hoc spinoffs, we will re-engineering their components into portable, generalized building blocks. We will address curation issues that arise from a service-oriented architecture (SOA), and the SOA life-cycle.
We will develop scalable extensions for our database cluster to deal with scaling large numerical simulations up to petabytes. We will turn these simulations into open numerical laboratories.
Our environment will allow users to interact with massive simulation datasets, run their own processing and simulations, use a shared "MyScratch" storage providing hundreds of TB of temporary storage to store results for subsequent analysis or sharing. Capabilities like this are currently unobtainable to the greater scientific community except for small subsets of researchers in specific domains. We can increase access to these capabilities across science community.
We will host additional scientific data sets within our integrated facility, expanding the scope of our scientific footprint. As we deploy and serve additional scientific domains via our building blocks, we will demonstrate the portability, generality, and economies of scale of the SciServer approach.
We will use our experience with the CasJobs Collaborative Environment to address the problem of small but complex data in the “Long Tail” of science, and in particular provide seamless integration across multiple datasets, including cross-queries with petabyte-scale managed datasets.
We will build upon our system foundations as we scale-up and take full ownership of the SDSS Archive (both database and flat files). We will ensure a scalable and robust environment for its continued operation, including supporting the SDSS4 project.
We will ensure that high value data, and the science that created it, is made widely available at an educational level by providing a framework to support new “Student Notebooks”.