Container applications in research computing and research data access
TimeWednesday, July 2511:15am - 12pm
DescriptionContainer-based technologies such as Docker, Singularity, and Shifter are emerging as a solution to a variety of problems in research computing and research data access. Containerization of research software stacks puts more control in the hands of the researcher to package software that can easily run in a variety of environments -- from laptops to cloud and high performance computing (HPC) systems. Containerized analysis environments can be easily run "near" very large research datasets, enabling exploratory access often unavailable in traditional research computing environments. In HPC systems, frameworks such as Singularity and Shifter reduce time required for installing and managing complex software stacks and potentially extend the viable lifetime of compute clusters. In cloud environments, containerized applications offer the opportunity for improved scalability and portability between cloud providers while simplifying traditional IT management tasks.
This panel will build on the 2017 Container Analysis Environments Workshop held at the National Center for Supercomputing Applications (NCSA) and supported by the National Data Service (NDS) and Data Exploration Lab (DXL). It will bring together researchers and research computing practitioners involved in the development of containerized solutions for research computing and data access to discuss applications of container technology in research environments -- from exploratory data analysis environments to HPC systems.