SciServerCompute: Bring Analysis Close to the Data

SciServer Compute is a powerful new resource for scientific research with big data - an environment for running data-intensive compute jobs online. SciServer Compute allows users to create Jupyter notebooks inside Docker containers co-located with a large institutional data repository at Johns Hopkins University. Our data center offers the complete dataset of the Sloan Digital Sky Survey (SDSS), as well as nearly a Petabyte of other data from many fields of science. This paper introduces the philosophy of SciServer Compute, outlines our Jupyter notebook cluster implementation, and presents the real-world science cases we have developed in collaboration with our scientific partners.